danvics/howdy
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge pull request #118 from boltgolt/dev

Browse source 
Version 2.5.0
This commit is contained in:
boltgolt 2019-01-06 16:44:57 +01:00 committed by GitHub
commit 40c290e2e7
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
34 changed files with 2826 additions and 329 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
.travis.yml
View file

@ -8,19 +8,16 @@ script:
# Install the binary, running the debian scripts in the process
- sudo apt install ../*.deb -y
# Confirm the cv2 module has been installed correctly
- sudo /usr/bin/env python3 -c "import cv2; print(cv2.__version__);"
# Confirm the face_recognition module has been installed correctly
- sudo /usr/bin/env python3 -c "import face_recognition; print(face_recognition.__version__);"
# Check if the username passthough works correctly with sudo
- 'howdy | ack-grep --passthru --color "current active user: travis"'
- 'sudo howdy | ack-grep --passthru --color "current active user: travis"'
# Go through function tests
- ./tests/importing.sh
- ./tests/passthrough.sh
# Skip PAM integration tests for now because of broken pamtester
# - ./tests/pam.sh
- ./tests/compare.sh
# Remove howdy from the installation
- sudo apt purge howdy -y
notifications:
email:
on_success: never
@ -34,3 +31,4 @@ addons:
- ack-grep
- devscripts
- fakeroot
- pamtester

8
README.md
View file

@ -28,6 +28,14 @@ Install the `howdy` package from the AUR. For AUR installation instructions, tak
You will need to do some additional configuration steps. Please read the [ArchWiki entry](https://wiki.archlinux.org/index.php/Howdy) for more information.
### Fedora
The `howdy` package is now available in a [Fedora COPR repository](https://copr.fedorainfracloud.org/coprs/luya/howdy/) by simply execute the following command from a terminal:
```
sudo dnf copr enable luya/howdy
sudo dnf install howdy
```
## Setup
After installation, you need to let Howdy learn your face. Run `sudo howdy add` to add a face model.

12
debian/changelog vendored
View file

@ -1,3 +1,15 @@
howdy (2.5.0) xenial; urgency=medium
* Added FFmpeg and v4l2 recorders (thanks @timwelch!)
* Added automatic PAM inclusion on installation
* Added optional notice on detection attempt (thanks @mrkmg!)
* Added support for grayscale frame encoding (thanks @dmig and @sapjunior!)
* Massively improved recognition speed (thanks @dmig!)
* Fixed typo in "timout" config value
* Removed unneeded dependencies (thanks @dmig!)
-- boltgolt <boltgolt@gmail.com> Sun, 06 Jan 2019 14:37:41 +0100
howdy (2.4.0) xenial; urgency=medium
* Cameras are now selected by path instead of by video device number (thanks @Rhiyo!)

4
debian/control vendored
View file

@ -9,7 +9,9 @@ Vcs-Git: https://github.com/boltgolt/howdy
Package: howdy
Homepage: https://github.com/boltgolt/howdy
Architecture: all
Depends: ${misc:Depends}, git, python3, python3-pip, python3-dev, python3-setuptools, libpam-python, fswebcam, libopencv-dev, python-opencv, cmake, streamer
Depends: ${misc:Depends}, curl|wget, python3, python3-pip, python3-dev, python3-setuptools, libpam-python, fswebcam, libopencv-dev, cmake, streamer
Recommends: libatlas-base-dev | libopenblas-dev | liblapack-dev
Suggests: nvidia-cuda-dev (>= 7.5)
Description: Howdy: Windows Hello style authentication for Linux.
Use your built-in IR emitters and camera in combination with face recognition
to prove who you are.

1
debian/install vendored
View file

@ -1,2 +1,3 @@
src/. lib/security/howdy
src/pam-config/. /usr/share/pam-configs
autocomplete/. usr/share/bash-completion/completions

227
debian/postinst vendored
View file

@ -2,6 +2,7 @@
# Installation script to install howdy
# Executed after primary apt install
def col(id):
"""Add color escape sequences"""
if id == 1: return "\033[32m"
@ -9,15 +10,15 @@ def col(id):
if id == 3: return "\033[31m"
return "\033[0m"
# Import required modules
import fileinput
import subprocess
import time
import sys
import os
import re
import signal
import fileinput
import urllib.parse
import tarfile
from shutil import rmtree, which
# Don't run unless we need to configure the install
# Will also happen on upgrade but we will catch that later on
@ -29,6 +30,7 @@ def log(text):
"""Print a nicely formatted line to stdout"""
print("\n>>> " + col(1) + text + col(0) + "\n")
def handleStatus(status):
"""Abort if a command fails"""
if (status != 0):
@ -36,6 +38,9 @@ def handleStatus(status):
sys.exit(1)
# Create shorthand for subprocess creation
sc = subprocess.call
# We're not in fresh configuration mode so don't continue the setup
if not os.path.exists("/tmp/howdy_picked_device"):
# Check if we have an older config we can restore
@ -53,15 +58,22 @@ if not os.path.exists("/tmp/howdy_picked_device"):
# Go through every setting in the old config and apply it to the new file
for section in oldConf.sections():
for (key, value) in oldConf.items(section):
# MIGRATION 2.3.1 -> 2.4.0
# If config is still using the old device_id parameter, convert it to a path
if key == "device_id":
key = "device_path"
value = "/dev/video" + value
# MIGRATION 2.4.0 -> 2.5.0
# Finally correct typo in "timout" config value
if key == "timout":
key = "timeout"
try:
newConf.set(section, key, value)
# Add a new section where needed
except configparser.NoSectionError as e:
except configparser.NoSectionError:
newConf.add_section(section)
newConf.set(section, key, value)
@ -69,6 +81,11 @@ if not os.path.exists("/tmp/howdy_picked_device"):
with open("/lib/security/howdy/config.ini", "w") as configfile:
newConf.write(configfile)
# Install dlib data files if needed
if not os.path.exists("/lib/security/howdy/dlib-data/shape_predictor_5_face_landmarks.dat"):
print("Attempting installation of missing data files")
handleStatus(subprocess.call(["./install.sh"], shell=True, cwd="/lib/security/howdy/dlib-data"))
sys.exit(0)
# Open the temporary file containing the device ID
@ -78,137 +95,139 @@ picked = in_file.read()
in_file.close()
# Remove the temporary file
subprocess.call(["rm /tmp/howdy_picked_device"], shell=True)
os.unlink("/tmp/howdy_picked_device")
log("Upgrading pip to the latest version")
# Update pip
handleStatus(subprocess.call(["pip3 install --upgrade pip"], shell=True))
handleStatus(sc(["pip3", "install", "--upgrade", "pip"]))
log("Cloning dlib")
log("Downloading and unpacking data files")
# Clone the dlib git to /tmp, but only the last commit
handleStatus(subprocess.call(["git", "clone", "--depth", "1", "https://github.com/davisking/dlib.git", "/tmp/dlib_clone"]))
# Run the bash script to download and unpack the .dat files needed
handleStatus(subprocess.call(["./install.sh"], shell=True, cwd="/lib/security/howdy/dlib-data"))
log("Downloading dlib")
dlib_archive = "/tmp/v19.16.tar.gz"
loader = which("wget")
LOADER_CMD = None
# If wget is installed, use that as the downloader
if loader:
LOADER_CMD = [loader, "--tries", "5", "--output-document"]
# Otherwise, fall back on curl
else:
loader = which("curl")
LOADER_CMD = [loader, "--retry", "5", "--location", "--output"]
# Assemble and execute the download command
cmd = LOADER_CMD + [dlib_archive, "https://github.com/davisking/dlib/archive/v19.16.tar.gz"]
handleStatus(sc(cmd))
# The folder containing the dlib source
DLIB_DIR = None
# A regex of all files to ignore while unpacking the archive
excludes = re.compile(
"davisking-dlib-\w+/(dlib/(http_client|java|matlab|test/)|"
"(docs|examples|python_examples)|"
"tools/(archive|convert_dlib_nets_to_caffe|htmlify|imglab|python/test|visual_studio_natvis))"
)
# Open the archive
with tarfile.open(dlib_archive) as tf:
for item in tf:
# Set the destenation dir if unset
if not DLIB_DIR:
DLIB_DIR = "/tmp/" + item.name
# extract only files sufficient for building
if not excludes.match(item.name):
tf.extract(item, "/tmp")
# Delete the downloaded archive
os.unlink(dlib_archive)
log("Building dlib")
# Start the build without GPU
handleStatus(subprocess.call(["cd /tmp/dlib_clone/; python3 setup.py install --yes USE_AVX_INSTRUCTIONS --no DLIB_USE_CUDA"], shell=True))
cmd = ["sudo", "python3", "setup.py", "install"]
cuda_used = False
flags = ""
# Get the CPU details
with open("/proc/cpuinfo") as info:
for line in info:
if "flags" in line:
flags = line
break
# Use the most efficient instruction set the CPU supports
if "avx" in flags:
cmd += ["--yes", "USE_AVX_INSTRUCTIONS"]
elif "sse4" in flags:
cmd += ["--yes", "USE_SSE4_INSTRUCTIONS"]
elif "sse3" in flags:
cmd += ["--yes", "USE_SSE3_INSTRUCTIONS"]
elif "sse2" in flags:
cmd += ["--yes", "USE_SSE2_INSTRUCTIONS"]
# Compile and link dlib
try:
sp = subprocess.Popen(cmd, cwd=DLIB_DIR, stdout=subprocess.PIPE)
except subprocess.CalledProcessError:
print("Error while building dlib")
raise
# Go through each line from stdout
while sp.poll() is None:
line = sp.stdout.readline().decode("utf-8")
if "DLIB WILL USE CUDA" in line:
cuda_used = True
print(line, end="")
log("Cleaning up dlib")
# Remove the no longer needed git clone
handleStatus(subprocess.call(["rm", "-rf", "/tmp/dlib_clone"]))
del sp
rmtree(DLIB_DIR)
print("Temporary dlib files removed")
log("Installing python dependencies")
log("Installing OpenCV")
# Install direct dependencies so pip does not freak out with the manual dlib install
handleStatus(subprocess.call(["pip3", "install", "--cache-dir", "/tmp/pip_howdy", "face_recognition_models==0.3.0", "Click>=6.0", "numpy", "Pillow"]))
log("Installing face_recognition")
# Install face_recognition though pip
handleStatus(subprocess.call(["pip3", "install", "--cache-dir", "/tmp/pip_howdy", "--no-deps", "face_recognition==1.2.2"]))
try:
import cv2
except Exception as e:
log("Reinstalling opencv2")
handleStatus(subprocess.call(["pip3", "install", "opencv-python"]))
handleStatus(subprocess.call(["pip3", "install", "--no-cache-dir", "opencv-python"]))
log("Configuring howdy")
# Manually change the camera id to the one picked
for line in fileinput.input(["/lib/security/howdy/config.ini"], inplace = 1):
print(line.replace("device_path = none", "device_path = " + picked), end="")
for line in fileinput.input(["/lib/security/howdy/config.ini"], inplace=1):
line = line.replace("device_path = none", "device_path = " + picked)
line = line.replace("use_cnn = false", "use_cnn = " + str(cuda_used).lower())
print(line, end="")
print("Camera ID saved")
# Secure the howdy folder
handleStatus(subprocess.call(["chmod 744 -R /lib/security/howdy/"], shell=True))
handleStatus(sc(["chmod 744 -R /lib/security/howdy/"], shell=True))
# Allow anyone to execute the python CLI
handleStatus(subprocess.call(["chmod 755 /lib/security/howdy"], shell=True))
handleStatus(subprocess.call(["chmod 755 /lib/security/howdy/cli.py"], shell=True))
handleStatus(subprocess.call(["chmod 755 -R /lib/security/howdy/cli"], shell=True))
os.chmod("/lib/security/howdy", 0o755)
os.chmod("/lib/security/howdy/cli.py", 0o755)
handleStatus(sc(["chmod 755 -R /lib/security/howdy/cli"], shell=True))
print("Permissions set")
# Make the CLI executable as howdy
handleStatus(subprocess.call(["ln -s /lib/security/howdy/cli.py /usr/local/bin/howdy"], shell=True))
handleStatus(subprocess.call(["chmod +x /usr/local/bin/howdy"], shell=True))
os.symlink("/lib/security/howdy/cli.py", "/usr/local/bin/howdy")
os.chmod("/usr/local/bin/howdy", 0o755)
print("Howdy command installed")
log("Adding howdy as PAM module")
# Will be filled with the actual output lines
outlines = []
# Will be fillled with lines that contain coloring
printlines = []
# Track if the new lines have been insterted yet
inserted = False
# Open the PAM config file
with open("/etc/pam.d/common-auth") as fp:
# Read the first line
line = fp.readline()
while line:
# Add the line to the output directly, we're not deleting anything
outlines.append(line)
# Print the comments in gray and don't insert into comments
if line[:1] == "#":
printlines.append("\033[37m" + line + "\033[0m")
else:
printlines.append(line)
# If it's not a comment and we haven't inserted yet
if not inserted:
# Set both the comment and the linking line
line_comment = "# Howdy IR face recognition\n"
line_link = "auth sufficient pam_python.so /lib/security/howdy/pam.py\n\n"
# Add them to the output without any markup
outlines.append(line_comment)
outlines.append(line_link)
# Make the print orange to make it clear what's being added
printlines.append("\033[33m" + line_comment + "\033[0m")
printlines.append("\033[33m" + line_link + "\033[0m")
# Mark as inserted
inserted = True
# Go to the next line
line = fp.readline()
# Print a file Header
print("\033[33m" + ">>> START OF /etc/pam.d/common-auth" + "\033[0m")
# Loop though all printing lines and use the enters from the file
for line in printlines:
print(line, end="")
# Print a footer
print("\033[33m" + ">>> END OF /etc/pam.d/common-auth" + "\033[0m" + "\n")
# Do not prompt for a yes if we're in no promt mode
if "HOWDY_NO_PROMPT" not in os.environ:
# Ask the user if this change is okay
print("Lines will be insterted in /etc/pam.d/common-auth as shown above")
ans = input("Apply this change? [y/N]: ")
# Abort the whole thing if it's not
if ans.lower().strip() != "y" or ans.lower().strip() == "yes":
print("Interpreting as a \"NO\", aborting")
sys.exit(1)
print("Adding lines to PAM\n")
# Write to PAM
common_auth = open("/etc/pam.d/common-auth", "w")
common_auth.write("".join(outlines))
common_auth.close()
# Activate the pam-config file
handleStatus(subprocess.call(["pam-auth-update --package"], shell=True))
# Sign off
print("Installation complete.")

3
debian/preinst vendored
View file

@ -9,6 +9,7 @@ def col(id):
if id == 3: return "\033[31m"
return "\033[0m"
import subprocess
import time
import sys
@ -24,7 +25,7 @@ if "upgrade" in sys.argv:
# Let the user know so he knows where to look on a failed install
print("Backup of Howdy config file created in /tmp/howdy_config_backup_v" + sys.argv[2] + ".ini")
except e:
except subprocess.CalledProcessError:
print("Could not make an backup of old Howdy config file")
# Don't continue setup when we're just upgrading

40
debian/prerm vendored
View file

@ -2,17 +2,11 @@
# Executed on deinstallation
# Completely remove howdy from the system
def col(id):
"""Add color escape sequences"""
if id == 1: return "\033[32m"
if id == 2: return "\033[33m"
if id == 3: return "\033[31m"
return "\033[0m"
# Import required modules
import subprocess
import sys
import os
from shutil import rmtree
# Only run when we actually want to remove
if "remove" not in sys.argv and "purge" not in sys.argv:
@ -24,24 +18,28 @@ if not os.path.exists("/lib/security/howdy/cli"):
# Remove files and symlinks
try:
subprocess.call(["rm /usr/local/bin/howdy"], shell=True)
except e:
os.unlink('/usr/local/bin/howdy')
except Exception:
print("Can't remove executable")
try:
subprocess.call(["rm /usr/share/bash-completion/completions/howdy"], shell=True)
except e:
os.unlink('/usr/share/bash-completion/completions/howdy')
except Exception:
print("Can't remove autocompletion script")
# Refresh and remove howdy from pam-config
try:
subprocess.call(["rm -rf /lib/security/howdy"], shell=True)
except e:
subprocess.call(["pam-auth-update --package"], shell=True)
subprocess.call(["rm /usr/share/pam-configs/howdy"], shell=True)
subprocess.call(["pam-auth-update --package"], shell=True)
except Exception:
print("Can't remove pam module")
# Remove full installation folder, just to be sure
try:
rmtree('/lib/security/howdy')
except Exception:
# This error is normal
pass
# Remove face_recognition and dlib
subprocess.call(["pip3 uninstall face_recognition face_recognition_models dlib -y --no-cache-dir"], shell=True)
# Print a tearbending message
print(col(2) + """
There are still lines in /etc/pam.d/common-auth that can't be removed automatically
Run "nano /etc/pam.d/common-auth" to remove them by hand\
""" + col(0))
# Remove dlib
subprocess.call(['pip3', 'uninstall', 'dlib', '-y', '--no-cache-dir'])

5
debian/source/options vendored
View file

@ -1,6 +1,7 @@
tar-ignore = ".git"
tar-ignore = "models"
tar-ignore = ".gitignore"
tar-ignore = ".github"
tar-ignore = "models"
tar-ignore = "tests"
tar-ignore = "README.md"
tar-ignore = "LICENSE"
tar-ignore = ".travis.yml"

37
src/cli.py
View file

@ -4,7 +4,6 @@
# Import required modules
import sys
import os
import subprocess
import getpass
import argparse
import builtins
@ -12,11 +11,11 @@ import builtins
# Try to get the original username (not "root") from shell
try:
user = os.getlogin()
except:
except Exception:
user = os.environ.get("SUDO_USER")
# If that fails, try to get the direct user
if user == "root" or user == None:
if user == "root" or user is None:
env_user = getpass.getuser().strip()
# If even that fails, error out
@ -28,37 +27,37 @@ if user == "root" or user == None:
# Basic command setup
parser = argparse.ArgumentParser(description="Command line interface for Howdy face authentication.",
formatter_class=argparse.RawDescriptionHelpFormatter,
add_help=False,
prog="howdy",
epilog="For support please visit\nhttps://github.com/boltgolt/howdy")
formatter_class=argparse.RawDescriptionHelpFormatter,
add_help=False,
prog="howdy",
epilog="For support please visit\nhttps://github.com/boltgolt/howdy")
# Add an argument for the command
parser.add_argument("command",
help="The command option to execute, can be one of the following: add, clear, config, disable, list, remove or test.",
metavar="command",
choices=["add", "clear", "config", "disable", "list", "remove", "test"])
help="The command option to execute, can be one of the following: add, clear, config, disable, list, remove or test.",
metavar="command",
choices=["add", "clear", "config", "disable", "list", "remove", "test"])
# Add an argument for the extra arguments of diable and remove
parser.add_argument("argument",
help="Either 0 (enable) or 1 (disable) for the disable command, or the model ID for the remove command.",
nargs="?")
help="Either 0 (enable) or 1 (disable) for the disable command, or the model ID for the remove command.",
nargs="?")
# Add the user flag
parser.add_argument("-U", "--user",
default=user,
help="Set the user account to use.")
default=user,
help="Set the user account to use.")
# Add the -y flag
parser.add_argument("-y",
help="Skip all questions.",
action="store_true")
help="Skip all questions.",
action="store_true")
# Overwrite the default help message so we can use a uppercase S
parser.add_argument("-h", "--help",
action="help",
default=argparse.SUPPRESS,
help="Show this help message and exit.")
action="help",
default=argparse.SUPPRESS,
help="Show this help message and exit.")
# If we only have 1 argument we print the help text
if len(sys.argv) < 2:

124
src/cli/add.py
View file

@ -1,33 +1,53 @@
# Save the face of the user in encoded form
# Import required modules
import subprocess
import time
import os
import sys
import json
import cv2
import configparser
import builtins
import cv2
import numpy as np
# Try to import face_recognition and give a nice error if we can't
# Try to import dlib and give a nice error if we can't
# Add should be the first point where import issues show up
try:
import face_recognition
import dlib
except ImportError as err:
print(err)
print("\nCan't import the face_recognition module, check the output of")
print("pip3 show face_recognition")
print("\nCan't import the dlib module, check the output of")
print("pip3 show dlib")
sys.exit(1)
# Get the absolute path to the current file
path = os.path.dirname(os.path.abspath(__file__))
# Get the absolute path to the current directory
path = os.path.abspath(__file__ + "/..")
# Test if at lest 1 of the data files is there and abort if it's not
if not os.path.isfile(path + "/../dlib-data/shape_predictor_5_face_landmarks.dat"):
print("Data files have not been downloaded, please run the following commands:")
print("\n\tcd " + os.path.realpath(path + "/../dlib-data"))
print("\tsudo ./install.sh\n")
sys.exit(1)
# Read config from disk
config = configparser.ConfigParser()
config.read(path + "/../config.ini")
if not os.path.exists(config.get("video", "device_path")):
print("Camera path is not configured correctly, please edit the 'device_path' config value.")
sys.exit(1)
use_cnn = config.getboolean("core", "use_cnn", fallback=False)
if use_cnn:
face_detector = dlib.cnn_face_detection_model_v1(path + "/../dlib-data/mmod_human_face_detector.dat")
else:
face_detector = dlib.get_frontal_face_detector()
pose_predictor = dlib.shape_predictor(path + "/../dlib-data/shape_predictor_5_face_landmarks.dat")
face_encoder = dlib.face_recognition_model_v1(path + "/../dlib-data/dlib_face_recognition_resnet_model_v1.dat")
user = builtins.howdy_user
# The permanent file to store the encoded model in
enc_file = path + "/../models/" + user + ".dat"
@ -47,8 +67,8 @@ except FileNotFoundError:
# Print a warning if too many encodings are being added
if len(encodings) > 3:
print("WARNING: Every additional model slows down the face recognition engine")
print("Press ctrl+C to cancel\n")
print("NOTICE: Each additional model slows down the face recognition engine slightly")
print("Press Ctrl+C to cancel\n")
print("Adding face model for the user " + user)
@ -56,15 +76,15 @@ print("Adding face model for the user " + user)
label = "Initial model"
# If models already exist, set that default label
if len(encodings) > 0:
if encodings:
label = "Model #" + str(len(encodings) + 1)
# Keep de default name if we can't ask questions
if builtins.howdy_args.y:
print("Using default label \"" + label + "\" because of -y flag")
print('Using default label "%s" because of -y flag' % (label, ))
else:
# Ask the user for a custom label
label_in = input("Enter a label for this new model [" + label + "]: ")
label_in = input("Enter a label for this new model [" + label + "] (max 24 characters): ")
# Set the custom label (if any) and limit it to 24 characters
if label_in != "":
@ -78,22 +98,35 @@ insert_model = {
"data": []
}
# Open the camera
video_capture = cv2.VideoCapture(config.get("video", "device_path"))
# Check if the user explicitly set ffmpeg as recorder
if config.get("video", "recording_plugin") == "ffmpeg":
# Set the capture source for ffmpeg
from recorders.ffmpeg_reader import ffmpeg_reader
video_capture = ffmpeg_reader(config.get("video", "device_path"), config.get("video", "device_format"))
elif config.get("video", "recording_plugin") == "pyv4l2":
# Set the capture source for pyv4l2
from recorders.pyv4l2_reader import pyv4l2_reader
video_capture = pyv4l2_reader(config.get("video", "device_path"), config.get("video", "device_format"))
else:
# Start video capture on the IR camera through OpenCV
video_capture = cv2.VideoCapture(config.get("video", "device_path"))
# Force MJPEG decoding if true
if config.get("video", "force_mjpeg") == "true":
if config.getboolean("video", "force_mjpeg", fallback=False):
# Set a magic number, will enable MJPEG but is badly documentated
video_capture.set(cv2.CAP_PROP_FOURCC, 1196444237)
# Set the frame width and height if requested
if int(config.get("video", "frame_width")) != -1:
video_capture.set(cv2.CAP_PROP_FRAME_WIDTH, int(config.get("video", "frame_width")))
fw = config.getint("video", "frame_width", fallback=-1)
fh = config.getint("video", "frame_height", fallback=-1)
if fw != -1:
video_capture.set(cv2.CAP_PROP_FRAME_WIDTH, fw)
if int(config.get("video", "frame_height")) != -1:
video_capture.set(cv2.CAP_PROP_FRAME_HEIGHT, int(config.get("video", "frame_height")))
if fh != -1:
video_capture.set(cv2.CAP_PROP_FRAME_HEIGHT, fh)
# Request a frame to wake the camera up
video_capture.read()
video_capture.grab()
print("\nPlease look straight into the camera")
@ -104,40 +137,53 @@ time.sleep(2)
enc = []
# Count the amount or read frames
frames = 0
dark_threshold = config.getfloat("video", "dark_threshold")
# Loop through frames till we hit a timeout
while frames < 60:
frames += 1
# Grab a single frame of video
# Don't remove ret, it doesn't work without it
ret, frame = video_capture.read()
gsframe = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
# Get the encodings in the frame
enc = face_recognition.face_encodings(frame)
# Create a histogram of the image with 8 values
hist = cv2.calcHist([gsframe], [0], None, [8], [0, 256])
# All values combined for percentage calculation
hist_total = np.sum(hist)
# If the image is fully black or the frame exceeds threshold,
# skip to the next frame
if hist_total == 0 or (hist[0] / hist_total * 100 > dark_threshold):
continue
frames += 1
# Get all faces from that frame as encodings
face_locations = face_detector(gsframe, 1)
# If we've found at least one, we can continue
if len(enc) > 0:
if face_locations:
break
# If 0 faces are detected we can't continue
if len(enc) == 0:
video_capture.release()
# If more than 1 faces are detected we can't know wich one belongs to the user
if len(face_locations) > 1:
print("Multiple faces detected, aborting")
sys.exit(1)
elif not face_locations:
print("No face detected, aborting")
sys.exit(1)
# If more than 1 faces are detected we can't know wich one belongs to the user
if len(enc) > 1:
print("Multiple faces detected, aborting")
sys.exit(1)
face_location = face_locations[0]
if use_cnn:
face_location = face_location.rect
# Totally clean array that can be exported as JSON
clean_enc = []
# Get the encodings in the frame
face_landmark = pose_predictor(frame, face_location)
face_encoding = np.array(face_encoder.compute_face_descriptor(frame, face_landmark, 1))
# Copy the values into a clean array so we can export it as JSON later on
for point in enc[0]:
clean_enc.append(point)
insert_model["data"].append(clean_enc)
insert_model["data"].append(face_encoding.tolist())
# Insert full object into the list
encodings.append(insert_model)

2
src/cli/config.py
View file

@ -17,4 +17,4 @@ elif "EDITOR" in os.environ:
editor = os.environ["EDITOR"]
# Open the editor as a subprocess and fork it
subprocess.call([editor, os.path.dirname(os.path.realpath(__file__)) + "/../config.ini"])
subprocess.call([editor, os.path.dirname(os.path.realpath(__file__)) + "/../config.ini"])

2
src/cli/disable.py
View file

@ -16,7 +16,7 @@ config = configparser.ConfigParser()
config.read(config_path)
# Check if enough arguments have been passed
if builtins.howdy_args.argument == None:
if builtins.howdy_args.argument is None:
print("Please add a 0 (enable) or a 1 (disable) as an argument")
sys.exit(1)

2
src/cli/list.py
View file

@ -8,7 +8,7 @@ import time
import builtins
# Get the absolute path and the username
path = os.path.dirname(os.path.realpath(__file__)) + "/.."
path = os.path.dirname(os.path.realpath(__file__)) + "/.."
user = builtins.howdy_user
# Check if the models file has been created yet

4
src/cli/remove.py
View file

@ -7,11 +7,11 @@ import json
import builtins
# Get the absolute path and the username
path = os.path.dirname(os.path.realpath(__file__)) + "/.."
path = os.path.dirname(os.path.realpath(__file__)) + "/.."
user = builtins.howdy_user
# Check if enough arguments have been passed
if builtins.howdy_args.argument == None:
if builtins.howdy_args.argument is None:
print("Please add the ID of the model you want to remove as an argument")
print("You can find the IDs by running:")
print("\n\thowdy list\n")

85
src/cli/test.py
View file

@ -1,14 +1,12 @@
# Show a windows with the video stream and testing information
# Import required modules
import face_recognition
import cv2
import configparser
import os
import sys
import json
import numpy
import time
import cv2
import dlib
# Get the absolute path to the current file
path = os.path.dirname(os.path.abspath(__file__))
@ -17,19 +15,27 @@ path = os.path.dirname(os.path.abspath(__file__))
config = configparser.ConfigParser()
config.read(path + "/../config.ini")
if config.get("video", "recording_plugin") != "opencv":
print("Howdy has been configured to use a recorder which doesn't support the test command yet")
print("Aborting")
sys.exit(12)
# Start capturing from the configured webcam
video_capture = cv2.VideoCapture(config.get("video", "device_path"))
# Force MJPEG decoding if true
if config.get("video", "force_mjpeg") == "true":
if config.getboolean("video", "force_mjpeg", fallback=False):
# Set a magic number, will enable MJPEG but is badly documented
video_capture.set(cv2.CAP_PROP_FOURCC, 1196444237)
# Set the frame width and height if requested
if int(config.get("video", "frame_width")) != -1:
video_capture.set(cv2.CAP_PROP_FRAME_WIDTH, int(config.get("video", "frame_width")))
fw = config.getint("video", "frame_width", fallback=-1)
fh = config.getint("video", "frame_height", fallback=-1)
if fw != -1:
video_capture.set(cv2.CAP_PROP_FRAME_WIDTH, fw)
if int(config.get("video", "frame_height")) != -1:
video_capture.set(cv2.CAP_PROP_FRAME_HEIGHT, int(config.get("video", "frame_height")))
if fh != -1:
video_capture.set(cv2.CAP_PROP_FRAME_HEIGHT, fh)
# Let the user know what's up
print("""
@ -39,6 +45,7 @@ Press ctrl+C in this terminal to quit
Click on the image to enable or disable slow mode
""")
def mouse(event, x, y, flags, param):
"""Handle mouse events"""
global slow_mode
@ -47,6 +54,21 @@ def mouse(event, x, y, flags, param):
if event == cv2.EVENT_LBUTTONDOWN:
slow_mode = not slow_mode
def print_text(line_number, text):
"""Print the status text by line number"""
cv2.putText(overlay, text, (10, height - 10 - (10 * line_number)), cv2.FONT_HERSHEY_SIMPLEX, .3, (0, 255, 0), 0, cv2.LINE_AA)
use_cnn = config.getboolean('core', 'use_cnn', fallback=False)
if use_cnn:
face_detector = dlib.cnn_face_detection_model_v1(
path + '/../dlib-data/mmod_human_face_detector.dat'
)
else:
face_detector = dlib.get_frontal_face_detector()
clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
# Open the window and attach a a mouse listener
cv2.namedWindow("Howdy Test")
cv2.setMouseCallback("Howdy Test", mouse)
@ -61,26 +83,31 @@ sec_frames = 0
fps = 0
# The current second we're counting
sec = int(time.time())
# recognition time
rec_tm = 0
# Wrap everything in an keyboard interupt handler
try:
while True:
# Inclement the frames
frame_tm = time.time()
# Increment the frames
total_frames += 1
sec_frames += 1
# Id we've entered a new second
if sec != int(time.time()):
if sec != int(frame_tm):
# Set the last seconds FPS
fps = sec_frames
# Set the new second and reset the counter
sec = int(time.time())
sec = int(frame_tm)
sec_frames = 0
# Grab a single frame of video
ret, frame = (video_capture.read())
ret, frame = video_capture.read()
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
frame = clahe.apply(frame)
# Make a frame to put overlays in
overlay = frame.copy()
@ -94,7 +121,7 @@ try:
# Fill with the overal containing percentage
hist_perc = []
# Loop though all values to calculate a pensentage and add it to the overlay
# Loop though all values to calculate a percentage and add it to the overlay
for index, value in enumerate(hist):
value_perc = float(value[0]) / hist_total * 100
hist_perc.append(value_perc)
@ -109,14 +136,11 @@ try:
# Draw a stripe indicating the dark threshold
cv2.rectangle(overlay, (8, 35), (20, 36), (255, 0, 0), thickness=cv2.FILLED)
def print_text(line_number, text):
"""Print the status text by line number"""
cv2.putText(overlay, text, (10, height - 10 - (10 * line_number)), cv2.FONT_HERSHEY_SIMPLEX, .3, (0, 255, 0), 0, cv2.LINE_AA)
# Print the statis in the bottom left
print_text(0, "RESOLUTION: " + str(height) + "x" + str(width))
print_text(1, "FPS: " + str(fps))
print_text(2, "FRAMES: " + str(total_frames))
print_text(0, "RESOLUTION: %dx%d" % (height, width))
print_text(1, "FPS: %d" % (fps, ))
print_text(2, "FRAMES: %d" % (total_frames, ))
print_text(3, "RECOGNITION: %dms" % (round(rec_tm * 1000), ))
# Show that slow mode is on, if it's on
if slow_mode:
@ -130,17 +154,22 @@ try:
# SHow that this is an active frame
cv2.putText(overlay, "SCAN FRAME", (width - 68, 16), cv2.FONT_HERSHEY_SIMPLEX, .3, (0, 255, 0), 0, cv2.LINE_AA)
rec_tm = time.time()
# Get the locations of all faces and their locations
face_locations = face_recognition.face_locations(frame)
face_locations = face_detector(frame, 1) # upsample 1 time
rec_tm = time.time() - rec_tm
# Loop though all faces and paint a circle around them
for loc in face_locations:
if use_cnn:
loc = loc.rect
# Get the center X and Y from the rectangular points
x = int((loc[1] - loc[3]) / 2) + loc[3]
y = int((loc[2] - loc[0]) / 2) + loc[0]
x = int((loc.right() - loc.left()) / 2) + loc.left()
y = int((loc.bottom() - loc.top()) / 2) + loc.top()
# Get the raduis from the with of the square
r = (loc[1] - loc[3]) / 2
r = (loc.right() - loc.left()) / 2
# Add 20% padding
r = int(r + (r * 0.2))
@ -158,9 +187,11 @@ try:
if cv2.waitKey(1) != -1:
raise KeyboardInterrupt()
frame_time = time.time() - frame_tm
# Delay the frame if slowmode is on
if slow_mode:
time.sleep(.55)
time.sleep(.5 - frame_time)
# On ctrl+C
except KeyboardInterrupt:

261
src/compare.py
View file

@ -5,33 +5,62 @@
import time
# Start timing
timings = [time.time()]
timings = {
"st": time.time()
}
# Import required modules
import cv2
import sys
import os
import json
import math
import configparser
import cv2
import dlib
import numpy as np
import _thread as thread
def init_detector(lock):
"""Start face detector, encoder and predictor in a new thread"""
global face_detector, pose_predictor, face_encoder
# Test if at lest 1 of the data files is there and abort if it's not
if not os.path.isfile(PATH + "/dlib-data/shape_predictor_5_face_landmarks.dat"):
print("Data files have not been downloaded, please run the following commands:")
print("\n\tcd " + PATH + "/dlib-data")
print("\tsudo ./install.sh\n")
lock.release()
sys.exit(1)
# Use the CNN detector if enabled
if use_cnn:
face_detector = dlib.cnn_face_detection_model_v1(PATH + "/dlib-data/mmod_human_face_detector.dat")
else:
face_detector = dlib.get_frontal_face_detector()
# Start the others regardless
pose_predictor = dlib.shape_predictor(PATH + "/dlib-data/shape_predictor_5_face_landmarks.dat")
face_encoder = dlib.face_recognition_model_v1(PATH + "/dlib-data/dlib_face_recognition_resnet_model_v1.dat")
# Note the time it took to initialize detectors
timings["ll"] = time.time() - timings["ll"]
lock.release()
# Read config from disk
config = configparser.ConfigParser()
config.read(os.path.dirname(os.path.abspath(__file__)) + "/config.ini")
def stop(status):
"""Stop the execution and close video stream"""
video_capture.release()
sys.exit(status)
# Make sure we were given an username to tast against
try:
if not isinstance(sys.argv[1], str):
sys.exit(1)
except IndexError:
sys.exit(1)
# The username of the authenticating user
# Make sure we were given an username to tast against
if len(sys.argv) < 2:
sys.exit(12)
# Get the absolute path to the current directory
PATH = os.path.abspath(__file__ + "/..")
# The username of the user being authenticated
user = sys.argv[1]
# The model file contents
models = []
@ -39,10 +68,19 @@ models = []
encodings = []
# Amount of ingnored dark frames
dark_tries = 0
# Total amount of frames captured
frames = 0
# face recognition/detection instances
face_detector = None
pose_predictor = None
face_encoder = None
# Try to load the face model from the models folder
try:
models = json.load(open(os.path.dirname(os.path.abspath(__file__)) + "/models/" + user + ".dat"))
models = json.load(open(PATH + "/models/" + user + ".dat"))
for model in models:
encodings += model["data"]
except FileNotFoundError:
sys.exit(10)
@ -50,127 +88,176 @@ except FileNotFoundError:
if len(models) < 1:
sys.exit(10)
# Put all models together into 1 array
for model in models:
encodings += model["data"]
# Read config from disk
config = configparser.ConfigParser()
config.read(PATH + "/config.ini")
# Add the time needed to start the script
timings.append(time.time())
# Get all config values needed
use_cnn = config.getboolean("core", "use_cnn", fallback=False)
timeout = config.getint("video", "timout", fallback=5)
dark_threshold = config.getfloat("video", "dark_threshold", fallback=50.0)
video_certainty = config.getfloat("video", "certainty", fallback=3.5) / 10
end_report = config.getboolean("debug", "end_report", fallback=False)
# Save the time needed to start the script
timings["in"] = time.time() - timings["st"]
# Import face recognition, takes some time
timings["ll"] = time.time()
# Start threading and wait for init to finish
lock = thread.allocate_lock()
lock.acquire()
thread.start_new_thread(init_detector, (lock, ))
# Start video capture on the IR camera
video_capture = cv2.VideoCapture(config.get("video", "device_path"))
timings["ic"] = time.time()
# Check if the user explicitly set ffmpeg as recorder
if config.get("video", "recording_plugin") == "ffmpeg":
# Set the capture source for ffmpeg
from recorders.ffmpeg_reader import ffmpeg_reader
video_capture = ffmpeg_reader(config.get("video", "device_path"), config.get("video", "device_format"))
elif config.get("video", "recording_plugin") == "pyv4l2":
# Set the capture source for pyv4l2
from recorders.pyv4l2_reader import pyv4l2_reader
video_capture = pyv4l2_reader(config.get("video", "device_path"), config.get("video", "device_format"))
else:
# Start video capture on the IR camera through OpenCV
video_capture = cv2.VideoCapture(config.get("video", "device_path"))
# Force MJPEG decoding if true
if config.get("video", "force_mjpeg") == "true":
# Set a magic number, will enable MJPEG but is badly documentated
if config.getboolean("video", "force_mjpeg", fallback=False):
# Set a magic number, will enable MJPEG but is badly documented
# 1196444237 is "GPJM" in ASCII
video_capture.set(cv2.CAP_PROP_FOURCC, 1196444237)
# Set the frame width and height if requested
if int(config.get("video", "frame_width")) != -1:
video_capture.set(cv2.CAP_PROP_FRAME_WIDTH, int(config.get("video", "frame_width")))
if int(config.get("video", "frame_height")) != -1:
video_capture.set(cv2.CAP_PROP_FRAME_HEIGHT, int(config.get("video", "frame_height")))
fw = config.getint("video", "frame_width", fallback=-1)
fh = config.getint("video", "frame_height", fallback=-1)
if fw != -1:
video_capture.set(cv2.CAP_PROP_FRAME_WIDTH, fw)
if fh != -1:
video_capture.set(cv2.CAP_PROP_FRAME_HEIGHT, fh)
# Capture a single frame so the camera becomes active
# This will let the camera adjust its light levels while we're importing for faster scanning
video_capture.read()
video_capture.grab()
# Note the time it took to open the camera
timings.append(time.time())
timings["ic"] = time.time() - timings["ic"]
# Import face recognition, takes some time
import face_recognition
timings.append(time.time())
# wait for thread to finish
lock.acquire()
lock.release()
del lock
# Fetch the max frame height
max_height = int(config.get("video", "max_height"))
max_height = config.getfloat("video", "max_height", fallback=0.0)
# Get the height of the image
height = video_capture.get(cv2.CAP_PROP_FRAME_HEIGHT) or 1
# Calculate the amount the image has to shrink
scaling_factor = (max_height / height) or 1
# Fetch config settings out of the loop
timeout = config.getint("video", "timeout")
dark_threshold = config.getfloat("video", "dark_threshold")
end_report = config.getboolean("debug", "end_report")
# Start the read loop
frames = 0
timings["fr"] = time.time()
while True:
# Increment the frame count every loop
frames += 1
# Stop if we've exceded the time limit
if time.time() - timings[3] > int(config.get("video", "timout")):
if time.time() - timings["fr"] > timeout:
stop(11)
# Grab a single frame of video
# Don't remove ret, it doesn't work without it
ret, frame = video_capture.read()
try:
# Convert from color to grayscale
# First processing of frame, so frame errors show up here
gsframe = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
except cv2.error:
print("\nUnknown camera, please check your 'device_path' config value.\n")
raise
# Create a histogram of the image with 8 values
hist = cv2.calcHist([frame], [0], None, [8], [0, 256])
hist = cv2.calcHist([gsframe], [0], None, [8], [0, 256])
# All values combined for percentage calculation
hist_total = int(sum(hist)[0])
hist_total = np.sum(hist)
# If the image is fully black, skip to the next frame
if hist_total == 0:
# If the image is fully black or the frame exceeds threshold,
# skip to the next frame
if hist_total == 0 or (hist[0] / hist_total * 100 > dark_threshold):
dark_tries += 1
continue
# Scrip the frame if it exceeds the threshold
if float(hist[0]) / hist_total * 100 > float(config.get("video", "dark_threshold")):
dark_tries += 1
continue
# Get the height and with of the image
height, width = frame.shape[:2]
# If the hight is too high
if max_height < height:
# Calculate the amount the image has to shrink
scaling_factor = max_height / float(height)
if scaling_factor != 1:
# Apply that factor to the frame
frame = cv2.resize(frame, None, fx=scaling_factor, fy=scaling_factor, interpolation=cv2.INTER_AREA)
# Save the new size for diagnostics
scale_height, scale_width = frame.shape[:2]
gsframe = cv2.resize(gsframe, None, fx=scaling_factor, fy=scaling_factor, interpolation=cv2.INTER_AREA)
# Get all faces from that frame as encodings
face_encodings = face_recognition.face_encodings(frame)
# Upsamples 1 time
face_locations = face_detector(gsframe, 1)
# Loop through each face
for face_encoding in face_encodings:
for fl in face_locations:
if use_cnn:
fl = fl.rect
# Fetch the faces in the image
face_landmark = pose_predictor(frame, fl)
face_encoding = np.array(face_encoder.compute_face_descriptor(frame, face_landmark, 1))
# Match this found face against a known face
matches = face_recognition.face_distance(encodings, face_encoding)
matches = np.linalg.norm(encodings - face_encoding, axis=1)
# Check if any match is certain enough to be the user we're looking for
match_index = 0
for match in matches:
match_index += 1
# Get best match
match_index = np.argmin(matches)
match = matches[match_index]
# Try to find a match that's confident enough
if match * 10 < float(config.get("video", "certainty")) and match > 0:
timings.append(time.time())
# Check if a match that's confident enough
if 0 < match < video_certainty:
timings["tt"] = time.time() - timings["st"]
timings["fr"] = time.time() - timings["fr"]
# If set to true in the config, print debug text
if config.get("debug", "end_report") == "true":
def print_timing(label, offset):
"""Helper function to print a timing from the list"""
print(" " + label + ": " + str(round((timings[1 + offset] - timings[offset]) * 1000)) + "ms")
# If set to true in the config, print debug text
if end_report:
def print_timing(label, k):
"""Helper function to print a timing from the list"""
print(" %s: %dms" % (label, round(timings[k] * 1000)))
print("Time spend")
print_timing("Starting up", 0)
print_timing("Opening the camera", 1)
print_timing("Importing face_recognition", 2)
print_timing("Searching for known face", 3)
# Print a nice timing report
print("Time spent")
print_timing("Starting up", "in")
print(" Open cam + load libs: %dms" % (round(max(timings["ll"], timings["ic"]) * 1000, )))
print_timing(" Opening the camera", "ic")
print_timing(" Importing recognition libs", "ll")
print_timing("Searching for known face", "fr")
print_timing("Total time", "tt")
print("\nResolution")
print(" Native: " + str(height) + "x" + str(width))
print(" Used: " + str(scale_height) + "x" + str(scale_width))
print("\nResolution")
width = video_capture.get(cv2.CAP_PROP_FRAME_WIDTH) or 1
print(" Native: %dx%d" % (height, width))
# Save the new size for diagnostics
scale_height, scale_width = frame.shape[:2]
print(" Used: %dx%d" % (scale_height, scale_width))
# Show the total number of frames and calculate the FPS by deviding it by the total scan time
print("\nFrames searched: " + str(frames) + " (" + str(round(float(frames) / (timings[4] - timings[3]), 2)) + " fps)")
print("Dark frames ignored: " + str(dark_tries))
print("Certainty of winning frame: " + str(round(match * 10, 3)))
# Show the total number of frames and calculate the FPS by deviding it by the total scan time
print("\nFrames searched: %d (%.2f fps)" % (frames, frames / timings["fr"]))
print("Dark frames ignored: %d " % (dark_tries, ))
print("Certainty of winning frame: %.3f" % (match * 10, ))
# Catch older 3-encoding models
if not match_index in models:
match_index = 0
print("Winning model: %d (\"%s\")" % (match_index, models[match_index]["label"]))
print("Winning model: " + str(match_index) + " (\"" + models[match_index]["label"] + "\")")
# End peacefully
stop(0)
# End peacefully
stop(0)

31
src/config.ini
View file

@ -2,6 +2,9 @@
# Press CTRL + X to save in the nano editor
[core]
# Print that face detection is being attempted
detection_notice = false
# Do not print anything when a face verification succeeds
no_confirmation = false
@ -14,33 +17,34 @@ ignore_ssh = true
# Auto dismiss lock screen on confirmation
# Will run loginctl unlock-sessions after every auth
# Expirimental, can behave incorrectly on some systems
# Experimental, can behave incorrectly on some systems
dismiss_lockscreen = false
# Disable howdy in the PAM
# The howdy command will still function
disabled = false
# Use CNN instead of HOG
# CNN model is much more accurate than the HOG based model, but takes much more
# computational power to run, and is meant to be executed on a GPU to attain reasonable speed.
use_cnn = false
[video]
# The certainty of the detected face belonging to the user of the account
# On a scale from 1 to 10, values above 5 are not recommended
certainty = 3.5
# The number of seconds to search before timing out
timout = 4
timeout = 4
# The path of the device to capture frames from
# Should be set automatically by the installer
# Should be set automatically by an installer if your distro has one
device_path = none
# Scale down the video feed to this maximum height
# Speeds up face recognition but can make it less precise
max_height = 320
# Force the use of Motion JPEG when decoding frames, fixes issues with
# YUYV raw frame deconding
force_mjpeg = false
# Set the camera input profile to this width and height
# The largest profile will be used if set to -1
# Automatically ignored if not a valid profile
@ -53,6 +57,19 @@ frame_height = -1
# The lower this setting is, the more dark frames are ignored
dark_threshold = 50
# The recorder to use. Can be either opencv (default), ffmpeg or pyv4l2.
# Switching from the default opencv to ffmpeg can help with grayscale issues.
recording_plugin = opencv
# Video format used by ffmpeg. Options include vfwcap or v4l2.
# FFMPEG only.
device_format = v4l2
# Force the use of Motion JPEG when decoding frames, fixes issues with YUYV
# raw frame decoding.
# OPENCV only.
force_mjpeg = false
[debug]
# Show a short but detailed diagnostic report in console
end_report = false

2
src/dlib-data/.gitignore vendored Normal file
View file

@ -0,0 +1,2 @@
*.dat
*.dat.bz2

7
src/dlib-data/Readme.md Normal file
View file

@ -0,0 +1,7 @@
Download and unpack `dlib` data files from https://github.com/davisking/dlib-models repository:
```shell
wget https://github.com/davisking/dlib-models/raw/master/dlib_face_recognition_resnet_model_v1.dat.bz2
wget https://github.com/davisking/dlib-models/raw/master/mmod_human_face_detector.dat.bz2
wget https://github.com/davisking/dlib-models/raw/master/shape_predictor_5_face_landmarks.dat.bz2
bunzip *bz2
```

25
src/dlib-data/install.sh Executable file
View file

@ -0,0 +1,25 @@
#!/bin/bash
echo "Downloading 3 required data files..."
# Check if wget is installed
if hash wget;then
# Check if wget supports the option to only show the progress bar
wget --help | grep -q "\--show-progress" && \
_PROGRESS_OPT="-q --show-progress" || _PROGRESS_OPT=""
# Download the archives
wget $_PROGRESS_OPT --tries 5 https://github.com/davisking/dlib-models/raw/master/dlib_face_recognition_resnet_model_v1.dat.bz2
wget $_PROGRESS_OPT --tries 5 https://github.com/davisking/dlib-models/raw/master/mmod_human_face_detector.dat.bz2
wget $_PROGRESS_OPT --tries 5 https://github.com/davisking/dlib-models/raw/master/shape_predictor_5_face_landmarks.dat.bz2
# Otherwise fall back on curl
else
curl --location --retry 5 --output dlib_face_recognition_resnet_model_v1.dat.bz2 https://github.com/davisking/dlib-models/raw/master/dlib_face_recognition_resnet_model_v1.dat.bz2
curl --location --retry 5 --output mmod_human_face_detector.dat.bz2 https://github.com/davisking/dlib-models/raw/master/mmod_human_face_detector.dat.bz2
curl --location --retry 5 --output shape_predictor_5_face_landmarks.dat.bz2 https://github.com/davisking/dlib-models/raw/master/shape_predictor_5_face_landmarks.dat.bz2
fi
# Uncompress the data files and delete the original archive
echo "Unpacking..."
bzip2 -d *.bz2

6
src/pam-config/howdy Normal file
View file

@ -0,0 +1,6 @@
Name: Howdy
Default: yes
Priority: 512
Auth-Type: Primary
Auth:
[success=end default=ignore] pam_python.so /lib/security/howdy/pam.py

26
src/pam.py
View file

@ -12,38 +12,46 @@ import ConfigParser
config = ConfigParser.ConfigParser()
config.read(os.path.dirname(os.path.abspath(__file__)) + "/config.ini")
def doAuth(pamh):
"""Starts authentication in a seperate process"""
# Abort is Howdy is disabled
if config.get("core", "disabled") == "true":
if config.getboolean("core", "disabled"):
sys.exit(0)
# Abort if we're in a remote SSH env
if config.get("core", "ignore_ssh") == "true":
if config.getboolean("core", "ignore_ssh"):
if "SSH_CONNECTION" in os.environ or "SSH_CLIENT" in os.environ or "SSHD_OPTS" in os.environ:
sys.exit(0)
# Alert the user that we are doing face detection
if config.get("core", "detection_notice") == "true":
pamh.conversation(pamh.Message(pamh.PAM_TEXT_INFO, "Attempting face detection"))
# Run compare as python3 subprocess to circumvent python version and import issues
status = subprocess.call(["/usr/bin/python3", os.path.dirname(os.path.abspath(__file__)) + "/compare.py", pamh.get_user()])
# Status 10 means we couldn't find any face models
if status == 10:
if config.get("core", "suppress_unknown") != "true":
if not config.getboolean("core", "suppress_unknown"):
pamh.conversation(pamh.Message(pamh.PAM_ERROR_MSG, "No face model known"))
return pamh.PAM_USER_UNKNOWN
# Status 11 means we exceded the maximum retry count
if status == 11:
elif status == 11:
pamh.conversation(pamh.Message(pamh.PAM_ERROR_MSG, "Face detection timeout reached"))
return pamh.PAM_AUTH_ERR
# Status 12 means we aborted
elif status == 12:
return pamh.PAM_AUTH_ERR
# Status 0 is a successful exit
if status == 0:
elif status == 0:
# Show the success message if it isn't suppressed
if config.get("core", "no_confirmation") != "true":
if not config.getboolean("core", "no_confirmation"):
pamh.conversation(pamh.Message(pamh.PAM_TEXT_INFO, "Identified face as " + pamh.get_user()))
# Try to dismiss the lock screen if enabled
if config.get("core", "dismiss_lockscreen") == "true":
if config.get("core", "dismiss_lockscreen"):
# Run it as root with a timeout of 1s, and never ask for a password through the UI
subprocess.Popen(["sudo", "timeout", "1", "loginctl", "unlock-sessions", "--no-ask-password"])
@ -53,18 +61,22 @@ def doAuth(pamh):
pamh.conversation(pamh.Message(pamh.PAM_ERROR_MSG, "Unknown error: " + str(status)))
return pamh.PAM_SYSTEM_ERR
def pam_sm_authenticate(pamh, flags, args):
"""Called by PAM when the user wants to authenticate, in sudo for example"""
return doAuth(pamh)
def pam_sm_open_session(pamh, flags, args):
"""Called when starting a session, such as su"""
return doAuth(pamh)
def pam_sm_close_session(pamh, flags, argv):
"""We don't need to clean anyting up at the end of a session, so returns true"""
return pamh.PAM_SUCCESS
def pam_sm_setcred(pamh, flags, argv):
"""We don't need set any credentials, so returns true"""
return pamh.PAM_SUCCESS

0
src/recorders/__init__.py Normal file
View file

132
src/recorders/ffmpeg_reader.py Normal file
View file

@ -0,0 +1,132 @@
# Class that simulates the functionality of opencv so howdy can use ffmpeg seamlessly
# Import required modules
import numpy
import sys
import re
from subprocess import Popen, PIPE
from cv2 import CAP_PROP_FRAME_WIDTH
from cv2 import CAP_PROP_FRAME_HEIGHT
try:
import ffmpeg
except ImportError:
print("Missing ffmpeg module, please run:")
print(" pip3 install ffmpeg-python\n")
sys.exit(12)
class ffmpeg_reader:
""" This class was created to look as similar to the openCV features used in Howdy as possible for overall code cleanliness. """
def __init__(self, device_path, device_format, numframes=10):
self.device_path = device_path
self.device_format = device_format
self.numframes = numframes
self.video = ()
self.num_frames_read = 0
self.height = 0
self.width = 0
self.init_camera = True
def set(self, prop, setting):
""" Setter method for height and width """
if prop == CAP_PROP_FRAME_WIDTH:
self.width = setting
elif prop == CAP_PROP_FRAME_HEIGHT:
self.height = setting
def get(self, prop):
""" Getter method for height and width """
if prop == CAP_PROP_FRAME_WIDTH:
return self.width
elif prop == CAP_PROP_FRAME_HEIGHT:
return self.height
def probe(self):
""" Probe the video device to get height and width info """
# Running this command on ffmpeg unfortunately returns with an exit code of 1, which is silly.
# Returns an error code of 1 and this text: "/dev/video2: Immediate exit requested"
args = ["ffmpeg", "-f", self.device_format, "-list_formats", "all", "-i", self.device_path]
process = Popen(args, stdout=PIPE, stderr=PIPE)
out, err = process.communicate()
return_code = process.poll()
# Worst case scenario, err will equal en empty byte string, b'', so probe will get set to [] here.
regex = re.compile(r"\s\d{3,4}x\d{3,4}")
probe = regex.findall(str(err.decode("utf-8")))
if not return_code == 1 or len(probe) < 1:
# Could not determine the resolution from ffmpeg call. Reverting to ffmpeg.probe()
probe = ffmpeg.probe(self.device_path)
height = probe["streams"][0]["height"]
width = probe["streams"][0]["width"]
else:
(height, width) = [x.strip() for x in probe[0].split("x")]
# Set height and width from probe if they haven't been set already
if height.isdigit() and self.get(CAP_PROP_FRAME_HEIGHT) == 0:
self.set(CAP_PROP_FRAME_HEIGHT, int(height))
if width.isdigit() and self.get(CAP_PROP_FRAME_WIDTH) == 0:
self.set(CAP_PROP_FRAME_WIDTH, int(width))
def record(self, numframes):
""" Record a video, saving it to self.video array for processing later """
# Eensure we have set our width and height before we record, otherwise our numpy call will fail
if self.get(CAP_PROP_FRAME_WIDTH) == 0 or self.get(CAP_PROP_FRAME_HEIGHT) == 0:
self.probe()
# Ensure num_frames_read is reset to 0
self.num_frames_read = 0
# Record a predetermined amount of frames from the camera
stream, ret = (
ffmpeg
.input(self.device_path, format=self.device_format)
.output("pipe:", format="rawvideo", pix_fmt="rgb24", vframes=numframes)
.run(capture_stdout=True, quiet=True)
)
self.video = (
numpy
.frombuffer(stream, numpy.uint8)
.reshape([-1, self.width, self.height, 3])
)
def read(self):
""" Read a sigle frame from the self.video array. Will record a video if array is empty. """
# First time we are called, we want to initialize the camera by probing it, to ensure we have height/width
# and then take numframes of video to fill the buffer for faster recognition.
if self.init_camera:
self.init_camera = False
self.video = ()
self.record(self.numframes)
return 0, self.video
# If we are called and self.video is empty, we should record self.numframes to fill the video buffer
if self.video == ():
self.record(self.numframes)
# If we've read max frames, but still are being requested to read more, we simply record another batch.
# Note, the video array is 0 based, so if numframes is 10, we must subtract 1 or run into an array index
# error.
if self.num_frames_read >= (self.numframes - 1):
self.record(self.numframes)
# Add one to num_frames_read. If we were at 0, that's fine as frame 0 is almost 100% going to be black
# as the IR lights aren't fully active yet anyways. Saves us one iteration in the while loop ni add/compare.py.
self.num_frames_read += 1
# Return a single frame of video
return 0, self.video[self.num_frames_read]
def release(self):
""" Empty our array. If we had a hold on the camera, we would give it back here. """
self.video = ()
self.num_frames_read = 0
def grab(self):
""" Redirect grab() to read() for compatibility """
self.read()

102
src/recorders/pyv4l2_reader.py Normal file
View file

@ -0,0 +1,102 @@
# Class that simulates the functionality of opencv so howdy can use v4l2 devices seamlessly
# Import required modules. lib4l-dev package is also required.
from recorders import v4l2
import fcntl
import numpy
import sys
from cv2 import cvtColor, COLOR_GRAY2BGR, CAP_PROP_FRAME_WIDTH, CAP_PROP_FRAME_HEIGHT
try:
from v4l2.frame import Frame
except ImportError:
print("Missing pyv4l2 module, please run:")
print(" pip3 install pyv4l2\n")
sys.exit(13)
class pyv4l2_reader:
""" This class was created to look as similar to the openCV features used in Howdy as possible for overall code cleanliness. """
# Init
def __init__(self, device_name, device_format):
self.device_name = device_name
self.device_format = device_format
self.height = 0
self.width = 0
self.probe()
self.frame = ""
def set(self, prop, setting):
""" Setter method for height and width """
if prop == CAP_PROP_FRAME_WIDTH:
self.width = setting
elif prop == CAP_PROP_FRAME_HEIGHT:
self.height = setting
def get(self, prop):
""" Getter method for height and width """
if prop == CAP_PROP_FRAME_WIDTH:
return self.width
elif prop == CAP_PROP_FRAME_HEIGHT:
return self.height
def probe(self):
""" Probe the video device to get height and width info """
vd = open(self.device_name, 'r')
fmt = v4l2.v4l2_format()
fmt.type = v4l2.V4L2_BUF_TYPE_VIDEO_CAPTURE
ret = fcntl.ioctl(vd, v4l2.VIDIOC_G_FMT, fmt)
vd.close()
if ret == 0:
height = fmt.fmt.pix.height
width = fmt.fmt.pix.width
else:
# Could not determine the resolution from ioctl call. Reverting to slower ffmpeg.probe() method
import ffmpeg
probe = ffmpeg.probe(self.device_name)
height = int(probe['streams'][0]['height'])
width = int(probe['streams'][0]['width'])
if self.get(CAP_PROP_FRAME_HEIGHT) == 0:
self.set(CAP_PROP_FRAME_HEIGHT, int(height))
if self.get(CAP_PROP_FRAME_WIDTH) == 0:
self.set(CAP_PROP_FRAME_WIDTH, int(width))
def record(self):
""" Start recording """
self.frame = Frame(self.device_name)
def grab(self):
""" Read a sigle frame from the IR camera. """
self.read()
def read(self):
""" Read a sigle frame from the IR camera. """
if not self.frame:
self.record()
# Grab a raw frame from the camera
frame_data = self.frame.get_frame()
# Convert the raw frame_date to a numpy array
img = (numpy.frombuffer(frame_data, numpy.uint8))
# Convert the numpy array to a proper grayscale image array
img_bgr = cvtColor(img, COLOR_GRAY2BGR)
# Convert the grayscale image array into a proper RGB style numpy array
img2 = (numpy.frombuffer(img_bgr, numpy.uint8).reshape([352, 352, 3]))
# Return a single frame of video
return 0, img2
def release(self):
""" Empty our array. If we had a hold on the camera, we would give it back here. """
self.video = ()
self.num_frames_read = 0
if self.frame:
self.frame.close()

1914
src/recorders/v4l2.py Normal file
View file

File diff suppressed because it is too large Load diff

29
tests/compare.sh Executable file
View file

@ -0,0 +1,29 @@
# TEST MODEL-FRAME COMPARE FUNCTIONS
set -o xtrace
set -e
# Make sure howdy is clean before starting
sudo howdy clear -y || true
# Learn match 1
sudo sed -i "s,device_path.*,device_path = $PWD\/tests\/video\/match1.m4v,g" /lib/security/howdy/config.ini
sudo howdy add -y
# Text compare matching with same camera input
sudo python3 /lib/security/howdy/compare.py $USER
# Change to match 2 and compare against the modal of match 1, which should fail
sudo sed -i "s,device_path.*,device_path = $PWD\/tests\/video\/match2.m4v,g" /lib/security/howdy/config.ini
! sudo python3 /lib/security/howdy/compare.py $USER
# Add match 2 as a model to compare both 1 and 2 at the same time
sudo howdy add -y
sudo python3 /lib/security/howdy/compare.py $USER
# Compare against a camera with no visible face
sudo sed -i "s,device_path.*,device_path = $PWD\/tests\/video\/noMatch.m4v,g" /lib/security/howdy/config.ini
! sudo python3 /lib/security/howdy/compare.py $USER
# Clean up
sudo howdy clear -y
sudo sed -i "s,device_path.*,device_path = none,g" /lib/security/howdy/config.ini

9
tests/importing.sh Executable file
View file

@ -0,0 +1,9 @@
# TEST INSTALLATION OF DEPENDENCIES
set -o xtrace
set -e
# Confirm the cv2 module has been installed correctly
sudo /usr/bin/env python3 -c "import cv2; print(cv2.__version__);"
# Confirm the dlib module has been installed correctly
sudo /usr/bin/env python3 -c "import dlib; print(dlib.__version__);"

32
tests/pam.sh Executable file
View file

@ -0,0 +1,32 @@
# TEST THE PAM INTEGRATION
set -o xtrace
set -e
# Make sure howdy is clean before starting
sudo howdy clear -y || true
# Change active camera to match video 1
sudo sed -i "s,device_path.*,device_path = $PWD/tests\/video\/match1.m4v,g" /lib/security/howdy/config.ini
# Let howdy add the match face
sudo howdy add -y
# Test the PAM auth
timeout 10 pamtester login $USER authenticate
# Clear the face models and change the camera to video 2
sudo howdy clear -y
sudo sed -i "s,device_path.*,device_path = $PWD\/tests\/video\/match2.m4v,g" /lib/security/howdy/config.ini
# Let howdy add the match face
sudo howdy add -y
# Try to open a elevated session through PAM
timeout 10 pamtester login $USER open_session
# Verify we can close sessions, even though howdy does not use this PAM function
timeout 10 pamtester login $USER close_session
# Clean up
sudo howdy clear -y
sudo sed -i "s,device_path.*,device_path = none,g" /lib/security/howdy/config.ini

7
tests/passthrough.sh Executable file
View file

@ -0,0 +1,7 @@
# TEST USER SUDO PASSTHOUGH (NON-ROOT)
set -o xtrace
set -e
# Check if the username passthough works correctly with sudo
howdy | ack-grep --passthru --color "current active user: travis"
sudo howdy | ack-grep --passthru --color "current active user: travis"

BIN
tests/video/match1.m4v Normal file
View file

Binary file not shown.

BIN
tests/video/match2.m4v Normal file
View file

Binary file not shown.

BIN
tests/video/noMatch.m4v Normal file
View file

Binary file not shown.