Footprint geometrical shape characterization method based on combination of polynomial fitting and right-angle filtering

Abstract

The invention provides a footprint geometrical shape characterization method based on combination of polynomial fitting and right-angle filtering, and the method comprises the steps: S1, obtaining a barefoot footprint image, and carrying out the preprocessing of the barefoot footprint image; s2, based on the pre-processed barefoot footprint image, judging a left barefoot and a right barefoot; s3, representing a left plantar leading edge line based on a 9-order polynomial; s4, representing a right plantar leading edge line based on a 9-order polynomial; s5, based on a right-angle filter, extracting toe contours; s6, extracting the outer contour of the barefoot footprint; and S7, representing the barefoot contour. According to the invention, a right-angle filter is adopted to find gullies between toes, namely concave and convex points, and the toes are separated through the concave and convex points between the toes; according to the method, the distance relation between the centers of the toes is considered, if the toes which are wrongly separated appear, the centers of all the connected domains can be solved, and if the distance between the center points of the two connected domains is too close, the two connected domains are expanded into one connected domain.

Description

technical field [0001] The present invention relates to the technical field of toe separation, in particular to a footprint geometry characterization method based on the combination of polynomial fitting and right-angle filtering. Background technique [0002] In the toe separation algorithm, there is a method based on region growth, which takes a plantar image as input, and after iterating through the snake algorithm, the outer contour shape of the foot region is obtained, and the sharp point of the outer contour shape is found, and the sharp point is used as the seed point , get the boundary area of ​​each toe through the method of region growth; then fit the anterior plantar edge line through the boundary area, select the middle point of each boundary area as the seed point, and then obtain the outline of each toe through the method of region growth, You can separate each toe. [0003] At present, the method of using region growing has achieved certain results, but this ...

AI Technical Summary

Problems solved by technology

[0003] At present, the method of using region growing has achieved certain results, but this method is suitable for plantar images. When barefoot footprint images are used as input, the selection of seed points will be difficult, and then it will not be possible to extract the front line of the plantar edge and separate the toes

Images