This project is an inplementation of adaboost. I developed this project on ubuntu 14.04. This project based on opencv-2.4.x which only used to decode image and some common image process just like resize.

Please be sure you in linux with opencv-2.4.x installed. If you want to work on other systems, you should rewrite the main function. Go into the source directory in terminal and type, make

there will generate three program in bin/ directory,

1. train used to train model
2. detect used to detect object
3. sample used to generate trainning samples
4. main no used

I prepared some shell scripts in bin directory

1. run.sh tranning model scripts
2. detect.sh detect object in image

First, you shoud prepare trainning samples and list positive samples' path in file pos_list.txt and list negative samples' path in file neg_list.txt. Then type ./run.sh

there will generate model file model.dat used to detect object. Then type ./detect.sh [image]

to detect object in image. It will show the result and write to out.jpg