An improved canny detection method for detecting human flexibility

Abstract Human flexibility is an important health indicator. In this chapter, a method for detecting human flexibility using image-processing technology is presented. There are two contributes to our work. Firstly, an improved Canny edge detection operator is proposed. The improved operator uses an adaptive Gaussian filter instead of the traditional Gaussian filter. Next, the 5 × 5 templates for calculating the gradient of the image is used with an increase of 45° and 135° in the direction of the template. Then, the threshold segmentation is realized by the maximum interclass variance method and the edge information of the image is obtained. Secondly, an algorithm for calculating the human feature point extraction required for the anteflexion angle of the human body is proposed. The experimental results of the human body anteflexion angle images collected show that the method proposed in this chapter is practical.