Intelligent recognition and counting detection method for circle-like objects for packaging industry

Abstract

The invention discloses an algorithm for intelligent recognition, counting, and detection of circle-like objects for the packaging industry. The effective area where the circle-like objects are located is obtained through images collected by industrial cameras; the image preprocessing method includes: using a median filter algorithm to remove original image noise, The binarized image after image denoising is obtained by using the adaptive local binary algorithm; the geometric feature extraction includes four detections, the first detection: detection of a circle-like object with a normal area after binarization; the second detection: detection of the second The defective circle-like objects remain after the first detection; the third detection: detect the circle-like objects in the uneven area; the fourth detection: detect the circle-like objects that may exist in the boundary. The invention can improve the counting accuracy of round-like three-dimensional objects, has low requirements on packaging conditions, and has strong detection robustness. At the same time, it can realize rapid and real-time fast online monitoring, and can be used for detection of various round-like objects.

Description

technical field [0001] The invention relates to a recognition technology detection algorithm, in particular to a packaging industry-oriented intelligent recognition counting detection method for circle-like objects. Background technique [0002] Circle-like objects are ubiquitous in life, such as rods, wine bottles, medicine bottles, cigarette boxes, etc. Such circle-like objects inevitably need statistics in life. At present, manual direct counting or no detection is often used for counting boxed round objects. Using manual direct detection counting method is labor-intensive, eyes are prone to fatigue, counting accuracy is low, and the error varies from person to person. If the computer technology with fast capability and modern machine vision technology are used to detect circle-like objects, it can not only reduce subjective interference, improve recognition accuracy, but also reduce the heavy labor burden of workers, thereby improving work efficiency. [0003] In the p...

AI Technical Summary

Problems solved by technology

[0004] Purpose of the invention: The purpose of the present invention is to solve the problems of high labor intensity, low work efficiency, low accuracy and error-prone in the prior art, and provide a solution to the problem of expensive special industrial cameras and supporting mobile devices, It solves the problem that the previous patent can only detect flat surface objects (concave-convex areas or edge interference cannot be solved), and the intelligent recognition and counting detection method for packaging industry-oriented circle-like objects

Images