This technical detail is my report time use PPT, although too brief, but still can grasp a lot of things, hereby share.
The relevant literature is the Integral Channel Features published in 2009 by Piotr Dollár and others.
The relevant literature translation can refer to the blog: http://write.blog.csdn.net/postedit/53965469
Related technologies can be consulted
Gaussian differential differentiate of gassion http://blog.csdn.net/abcjennifer/article/details/7639488 http://blog.csdn.net/ majinlei121/article/details/51240591 grayscale, grayscale, grayscale image http://www.cnblogs.com/heartchord/p/5903011.html integral image (Integral Image) and Integral histogram (Integral histogram) The principle and derivation of http://blog.sina.com.cn/s/blog_5562b0440102wgxs.html Adaboost algorithm/HTTP/ blog.csdn.net/u012258999/article/details/42457577 some notes on cascade discrete adaboost (sample normalization, weak classifier, strong classifier threshold, cascade parameter setting)/HTTP// Blog.sina.com.cn/s/blog_78fd98af01010zy6.html target recognition based on hog sliding window http://blog.csdn.net/orsinozhu/article/details/ 40554211 Multi-scale sliding window Muti-scale http://blog.csdn.net/fclovw/article/details/52421598
"Integral Channel Features technical details" below