Contact nameAshutosh Morde
Contact emailashu@intuvisiontech.com
Contact university/companyintuVision Inc.
Method's nameChebyshev probability approach
ReferenceA. Morde, X. Ma, S. Guler "Learning a background model for change detection", in proc of IEEE Workshop on Change Detection, 2012 [IntuVision]
Processing time~24fps on 1920x1080 video. C++/CUDA on Core i7-2600 CPU @ 3.4Ghz; NVIDIA GeForce GT 545
Web pagehttp://www.intuvisiontech.com/products/panoptes.php
Parameters motion-sensitivity=0.5, min-bg-pct=0.82, image-q-len=5, update-percent=0.1,filter-noise level="auto",shadow-removal = enabled

You can download the change detection results of this method by clicking here.

Video TP FP FN TN Nb Shadow Error
overpass 1851642 282398 126250 145360363 5059
canoe 928708 85289 114661 28334432 0
fountain01 48632 271545 28871 93011510 0
fountain02 237884 19176 29345 123943155 0
fall 14979729 1640549 3358268 1015149358 259506
boats 1160339 258675 390512 245460944 228
Video TP FP FN TN Nb Shadow Error
copyMachine 62085568 16739816 6648097 906562726 0
bungalows 4847456 1779634 309332 79060084 1739790
busStation 2452874 681656 498335 76333846 210847
peopleInShade 4435164 1126487 470192 80874173 713558
backdoor 1606212 27253 418385 99800534 0
cubicle 8605192 2423656 1770001 515479119 83463
Video TP FP FN TN Nb Shadow Error
lakeSide 412250 104137 7560831 408054277 0
park 504772 134534 210365 34360606 0
diningRoom 17153273 695901 2311024 206488366 0
corridor 11018334 610603 1321823 359376422 0
library 58123272 2511005 3837993 256969510 0