Fall detection based on depth images via wavelet moment
Yupeng Ding, Hongjun Li, Chaobo Li, Ke Xu, Pengzhen Guo · 2017
Fall detection plays an important role in the detection of human abnormal behaviors. In this paper, a fall detection algorithm based on depth images via wavelet moment is proposed. Firstly, we normalize the image according to each pixel in the image relative to the distance from the centroid, then polar coordinate the normalized image. Secondly, Fast Fourier Transform (FFT) of the picture is performed. Thirdly, the feature vectors of the image are extracted by using wavelet transform. Finally, using the minimum distance and Support Vector Machine (SVM) classification methods recognizes human behaviors. Numerous experiments are carried out on a large number of human behavior samples and the averaged success rate of this algorithm is more than 90%. The experimental results show that the proposed algorithm is robust and has good detecting ability and application prospect.