Patch position error detection method based on digital image

Abstract

The invention discloses a patch position error detection method based on a digital image. The method comprises the following steps: 1) denoising an image by adopting a weighted mean filter, improvingthe image contrast by adopting image histogram equalization, and extracting image edge features by adopting an improved Canny algorithm; 2) intercepting a template and recording a matching standard point, and carrying out template matching by adopting an improved SAD algorithm (absolute error sum) according to the result of the step 1); and 3) finally, finding a matching standard point by applyingthe matching standard point offset vector, and further calculating the patch position error based on the digital image. According to the patch position error detection method based on the digital image, patch edge features and position features based on the digital image are fully combined, and patch position error detection can be completed quickly and accurately.

Description

technical field [0001] The invention relates to a position error detection method, in particular to a digital image-based patch position error detection method. Background technique [0002] In the process of producing optical fiber transceiver PCB boards, it is necessary to use an automatic placement machine for dispensing and placement on the PCB board. In actual production, due to the high positioning accuracy of product placement, the traditional automatic placement machine cannot guarantee that the placement accuracy meets the manufacturer's requirements. Therefore, after the chip placement process, the chip position detection must be carried out, and the traditional chip position detection mainly relies on manual operation detection, which not only has low work efficiency, but also has large detection errors, so the manual detection method can no longer meet the large-scale production of optical fiber transceiver PCB boards. Require. [0003] The traditional feature ...

AI Technical Summary

Problems solved by technology

Among them, the classic Canny feature extraction method needs to calculate the gradient direction value of all points on the non-maximum value suppression algorithm, and perform interpolation operation to judge whether the point is the maximum value of the gradient direction; while the SAD algorithm (absolute error sum) needs to accumulate each The sum of the absolute value error between the gradient value and the template, the time cost of the two is relatively large, and it is not suitable for the rapid production requirements of industrial automation

Images