Overlapped particulate matter layered counting method based on a color image and a depth image

Abstract

The invention provides an overlapped particulate matter layered counting method based on a color image and a depth image. Color and depth images are acquired by the Kinect camera and registered, and improved K-is adopted. a color image target area is extracted by using a means algorithm, dividing the color image target area into single target particles, adhesion types and overlapping types by using contour characteristics of the color image target area, and reconstructing an extractable contour by using an interpolation algorithm to obtain a corresponding area target quantity; For overlapped target contours which cannot be extracted by the method, the average area of a single particulate matter can be estimated, and the corresponding number is obtained by dividing the total area of the target area by the average area; And for the upper layer target which cannot be extracted by using the color image, threshold segmentation is performed on the depth image of the upper layer target so asto complete extraction and classification counting of the upper layer target. According to the method, the paper-skin walnuts which are uniform in size and similar to a circle are adopted as researchobjects, the walnuts which are placed at will are counted, the average correct rate reaches 99.38%, it is indicated that the method is effective, and meanwhile a new idea is provided for counting overlapped particulate matter.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, and relates to the collection and processing of depth images, in particular to classification identification and hierarchical counting of overlapping particles based on color images and depth images. Background technique [0002] In recent years, the automatic counting method using image processing technology has been applied in many fields, especially in the field of agricultural automation. The counting method based on machine vision has effectively overcome many shortcomings of manual counting. In the process of particle counting, there will be multiple overlapping occlusions, which seriously affects the accuracy of counting. Therefore, the counting of overlapping particles has become a difficult problem in automatic counting systems, and has also attracted the attention of many scholars. Aiming at the research status of the current automatic counting method of overlapping par...

AI Technical Summary

Problems solved by technology

In the process of particle counting, there will be multiple overlapping occlusions, which seriously affects the accuracy of counting. Therefore, the counting of overlapping particles has become a difficult problem in automatic counting systems, and has also attracted the attention of many scholars.

Aiming at the research status of the current automatic counting method of overlapping particles, the counting accuracy is not high and the counting cannot be counted when the particles are completely overlapped, this paper proposes a layered counting method for overlapping particles

Images