A multi-target counting method based on an image recognition technology

Abstract

The invention relates to a multi-target counting method based on an image recognition technology, and the method comprises the following steps: obtaining an original image, and obtaining a target image through recording the target image; Image preprocessing; Current objects are independently screened and adhered; morphological operation is carried out on the binary image; Edge detection algorithmis adopted, a continuous image edge is identified; according to the area parameter of the target object,The image is divided, the edge smaller than the given area threshold value can be regarded as anindependent target object, the edge larger than the given area threshold value can be regarded as a possible adhesion target object, and therefore non-adhesion and adhesion area distinguishing is achieved, statistics is conducted on the non-adhesion area, and a strong adhesion area is processed. According to the multi-target counting method based on the image recognition technology and the adhesion region division method, the filling operation after the morphological corrosion operation is carried out is carried out in the mother image, the uniqueness of the area threshold value is ensured, and the subsequent operation process is greatly simplified.

Description

technical field [0001] The invention belongs to the technical fields of image recognition and machine vision, and in particular relates to a multi-target counting method based on image recognition technology. Background technique [0002] With the development of computer technology, image recognition technology has been widely used in production and life; the first category is various image recognition in general scenarios (the targets to be recognized are fixed and the types are limited), mainly for text, license plate, Fingerprint identification; the other type is identification in special scenarios, for example: in the fields of industrial automation, product sorting on assembly lines, and medical image screening (the types of targets that need to be identified are not fixed). The implementation principle of the former: for the first type of general application scenarios, the current mainstream method mainly uses neural network algorithms to classify and compare samples. ...

AI Technical Summary

Problems solved by technology

[0003] For the second type of application scenarios, most of them focus on the extraction and recognition of single target object features or defect detection, medical cell counting, etc. First, a large number of samples need to be trained, and second, over-fitting is prone to occur, and the range of sample sets Inner images can be accurately counted and identified, but for unknown images outside the sample set, the recognition error rate is high, which cannot meet the needs of practical applications; for accurate identification and counting

Images