An Android People Counting System Based On OpenCV

• Main Authors: Chih-Chun Peng, 彭治群
• Other Authors: Jia-Yin Wang
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/90718613829494355725

碩士 === 中原大學 === 電機工程研究所 === 105 === In this research, a people counting system is implemented based on an Android smart phone and the face detection functions of OpenCV. The face detection method is based on Integral Image, Adaboost and Cascade of Classifier algorithms, together with the Harr-like face classifier provided by OpenCV. The proposed system can be used to count the people number in a picture which may be taken by the smart phone or selected from a gallery, and can be executed in a continuous real-time detection scenario. We also compare the detection rates of OpenCV and another software Dlib based on 100 selected pictures.