392 commits

This branch

This branch

All branches

Author	SHA1	Message	Date
boltgolt	40c290e2e7	Merge pull request #118 from boltgolt/dev ... Version 2.5.0	2019-01-06 16:44:57 +01:00
boltgolt	e8d8692f40	Typo fix	2019-01-06 16:32:18 +01:00
boltgolt	c3649dca3a	Added changelog	2019-01-06 16:04:57 +01:00
boltgolt	77fd423df9	Added upgrade step and fixed lintian errors	2019-01-06 15:36:03 +01:00
boltgolt	3c8db931b9	Moved recorders to new folder	2019-01-06 15:08:40 +01:00
boltgolt	6e1085110b	Merge pull request #117 from mrkmg/add-checking-notification ... Added notice that facial detection is occurring	2019-01-05 01:11:45 +01:00
Kevin Gravier	de6dac1fd1	Added config option to enable/disable notification ... Default is off	2019-01-04 15:26:38 -05:00
boltgolt	133f9a5ccb	Reverting dlib install change because it did very little to speed up tests ... This reverts commit 1714579e6e.	2019-01-03 15:05:53 +01:00
boltgolt	1714579e6e	Removing dlib from travis for faster tests	2019-01-03 14:53:59 +01:00
boltgolt	a4928144cc	Rewrote download script, fixed tests and other misc changes	2019-01-03 14:35:07 +01:00
boltgolt	6928c22b4e	Merge pull request #116 from timwelch/dev ... Adding support for v4l2 devices directly in python	2019-01-02 23:48:12 +01:00
boltgolt	2917e8ada4	Code cleanup	2019-01-02 23:36:09 +01:00
boltgolt	e38422c8a0	Installer fixes	2019-01-02 22:57:00 +01:00
Kevin	140c8db081	Added notice that facial detection is occuring ... Added a pam conversation message that the facial detection is being performed.	2019-01-02 13:56:02 -05:00
Tim Welch	6a73634ac2	Adding support for v4l2 devices directly in python ... v4l2l.py is included in src dir. The version you get from pip install does not work with python3 (at least not on my system), so I had to modify a bunch of the functions returning a list() of range() instead of the original range().	2018-12-23 12:08:18 -05:00
boltgolt	54998e9a97	Merge branch 'dmig-master' into dev	2018-12-21 16:52:13 +01:00
boltgolt	0db3cb3da2	Merge branch 'master' of https://github.com/dmig/howdy into dmig-master	2018-12-21 16:51:23 +01:00
Dmitriy Geels	b2f30d13c3	Merge branch 'master' into master	2018-12-21 17:10:14 +03:00
boltgolt	f4db5ee94b	Formatting fixes	2018-12-13 20:53:18 +01:00
boltgolt	f8cfde6f72	Fixed a few small things from PR #109	2018-12-13 20:10:08 +01:00
boltgolt	faa3e28563	Merge pull request #109 from timwelch/dev ... Initial FFMPEG support for HP Spectre x360 13t IR camera	2018-12-13 18:58:31 +01:00
dmig	a8d68d945b	no changes, comment	2018-12-12 16:43:27 +07:00
dmig	443c5e7063	proper library load timing; add total time reporting	2018-12-12 16:30:52 +07:00
dmig	cc1b9785a8	replace threading module with low-level _thread; leave only one initialization helper thread	2018-12-12 15:04:11 +07:00
dmig	5fcb7383ca	reorder lines for readability	2018-12-12 15:00:59 +07:00
dmig	01c6364e65	simplify argument check; comments typos	2018-12-12 14:58:03 +07:00
dmig	7aac419bd7	importing all modules at the beginning is ~40% faster	2018-12-12 14:45:23 +07:00
Tim Welch	39daaa272b	pretty up the code a bit for ffmpeg resolution detection	2018-12-11 20:53:42 -05:00
Tim Welch	546015d973	using ffmpeg to grab resolution to save 700ms on startup	2018-12-11 19:00:55 -05:00
dmig	4f11da686f	one more detector init thread	2018-12-12 00:15:35 +07:00
dmig	62aabdf480	initialize recognizers in threads instead of camera init	2018-12-12 00:08:13 +07:00
dmig	f4d231cca8	proper cam init time measuring	2018-12-11 23:44:30 +07:00
dmig	c90d977352	fix broken previous commit	2018-12-11 16:40:06 +07:00
dmig	bedfa8020c	more informative debug	2018-12-11 16:30:45 +07:00
dmig	02a9aee63a	convert timings from list to dict; move camera initialization to separate thread	2018-12-11 16:24:35 +07:00
Tim Welch	693eef964a	Initial ffmpeg support ... Cleaned up unused imports and hard coded width/height values.	2018-12-09 12:12:50 -05:00
Tim Welch	2013338155	Add initial support for ffmpeg ... Fixes HP Spectre x360 13t IR camera issue. The IR camera is detected by OpenCV but fails to work properly. FFMPEG works using /dev/video2 device. Added an ffmpeg class that mimics the behavior and currently used functions of opencv.	2018-12-09 12:06:31 -05:00
boltgolt	1f0ed1267a	The dyslexia fix	2018-12-09 12:47:29 +01:00
dmig	d1d4d0af23	remove last face_recognition references	2018-12-09 14:28:16 +07:00
dmig	2c48350b23	replace face_recognition with dlib+numpy; add video_capture.release(); minor code cleanup	2018-12-09 14:27:21 +07:00
dmig	8671be425d	use defaults for config values; simplify path detection	2018-12-09 14:24:50 +07:00
dmig	b465982092	dark frame detection	2018-12-09 14:22:55 +07:00
dmig	377241bd2e	don't frighten user with slowdown message, it's not so significant; notify user about max name length	2018-12-09 13:50:35 +07:00
dmig	ff02a723d4	replace most subprocess calls with native calls; download and install dlib from tarball instead of making git clone; configure dlib according to cpu flags; do not install opencv using pip; set use_cnn option automatically; minor code cleanup	2018-12-09 13:30:37 +07:00
dmig	9c2507ed13	replace subprocess calls with native calls	2018-12-09 13:25:02 +07:00
dmig	82e4e9e10f	replace git dependency with wget|curl; fix opencv dependency; add recommended and suggested packages	2018-12-09 13:23:55 +07:00
dmig	1b8a7dc449	replace face_recognition with dlib+numpy; support for cnn detector; calculate scaling_factor outside of main cycle; use grayscale image for darkframe and face detection; fix winning model index detection; some minor code cleanups	2018-12-09 13:21:44 +07:00
dmig	c8c481aed2	use PATH constant	2018-12-09 13:15:31 +07:00
dmig	01e706fce6	speedup startup a bit	2018-12-09 13:12:28 +07:00
dmig	29c3a31a5b	fix models loading problem	2018-12-09 13:11:08 +07:00