Study And Implementation Of Face Detection Based On Adaboost Algorithm

Face detection in pattern recognition and machine vision technology as a long-term subject of concern, has high academic value and commercial value. With the rapid development of intelligent computing technology, new methods and new technologies will continue to be used, which bring more vitality to the research, and detection result will be increasingly better and faster.The research started from the adaboost algorithm, which was optimized from the training process and the detection process, and on this basis face detection and tracking system was developed.Firstly, basing on the classical principles of the adaboost algorithm, the training process and the detection process were achieved respectively and independently. Secondly, in view of the training process consumed such a long time, the number of features and the steps of the training process were optimized respectively, so that training time was reduced by nearly 50%. Then, from the point of speed required by real-time detection system, a face tracking algorithm based on the adaboost algorithm was proposed, and the average processing time of single frame was shortened sharply. Finally, after realizing the detection and tracking algorithm, the face detection and tracking system was developed, including image-based face detection subsystem, video-based face detection and tracking subsystem and video surveillance subsystem. In addition, by embedded of the complexion module and the principal component analysis module, the performance of the system was greatly improved.Because of realizing the adaboost algorithm independently, the relevant procedures could be transplanted easily, and a good foundation was lay for further study and improving the application.