Method for detecting component size based on matching pursuit (MP) wavelet filtering and detecting system thereof

Abstract

The invention discloses a method for detecting component size based on matching pursuit (MP) wavelet filtering. The method comprises the steps of (1), calibrating a charge coupled device (CCD) camera; (2), acquiring a component image and performing MP wavelet filtering on an image; (3), performing differential operator edge detection on the image; (4) performing sub-pixel edge detection on an edge pixel; and (5) calculating the size of the component according to a sub-pixel coordinate to generate computer-aided design (CAD) graph. According to the method, the image of the component is subjected to the MP wavelet filtering, the filtered image is subjected to the sub-pixel edge detection, and the size of the component is calculated and extracted according to the sub-pixel coordinate, so that the detection accuracy can be correspondingly improved, and the detection speed is quickened. The invention discloses a detecting system based on the detecting method. The system comprises an image acquisition unit, an image processing unit and an image generating unit. Because the camera is combined with a personal computer (PC), image acquisition and image processing on components on the production line can be realized, so that the timeliness for detecting the components is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of detection, and in particular relates to a part size detection method and a detection system based on MP (maximum algebra) wavelet filtering. Background technique [0002] Parts are widely used in military industry and civilian use, so it is of great significance to accurately detect the size of parts. The detection of part size requires high detection accuracy to reduce errors and improve product quality. The size detection of parts on the assembly line requires a higher detection speed to improve the detection efficiency. Therefore, in order to efficiently detect the size of parts, it is necessary to seriously consider the detection speed and accuracy. [0003] There are mainly the following ways to detect the size of parts on the existing assembly line: (1) manual detection, the operator uses the detection tool to sample the parts, the detection accuracy of this detection method depends on the detecti...

AI Technical Summary

Problems solved by technology

The existing image filtering methods for parts are often difficult to balance the speed and accuracy requirements: mean filtering is easy to cause image blur, median filtering is not ideal for noise filtering of non-salt and pepper noise, and the actual noise is often not salt and pepper noise, so it is difficult to meet Accuracy requirements; wavelet filtering has a good filtering effect, but it takes a long time and it is difficult to meet the speed requirements

Existing edge detection methods for part images use common differential operator edge detection methods: Robert edge detection, Prewitt edge detection, Sobel edge detection, Laplacian edge detection and other methods. Although these edge detection methods have faster detection speeds, they are difficult Meet the detection accuracy, especially for parts with high requirements for detection accuracy; sub-pixel edge detection based on moments and sub-pixel edge detection based on curve fitting have high detection accuracy, but it is difficult to meet the detection speed requirements; based on 4 or The sub-pixel edge detection method of Sobel edge detection in 8 directions combined with interpolation method. Since the Sobel operator is 3×3 in size, the calculation amount is large, and it is difficult to meet the detection speed requirements.

[0004] In addition, the existing part size detection does not have the AutoCAD generation and display function, so the edge and size image of the part cannot be seen intuitively, which brings great benefits to the subsequent part processing, such as determining the defect position of the defective part and repairing the defective part. Great inconvenience, resulting in a waste of resources; and the AutoCAD format is convenient for the standardized storage format and processing of data

Images