FPC defect detection method based on improved MASK RCNN

Abstract

The invention discloses an FPC defect detection method based on an improved MASK RCNN, and the method comprises the following steps: S1, carrying out the ROI processing and image cutting, collecting a target detection image, carrying out the data preprocessing of an FPC original image, and cutting the data into small images suitable for network input; S2, performing data enhancement processing on the preprocessed image, and expanding a data set of model training; S3, carrying out manual labeling; S4, determining the transverse stacking times N of the block of the feature map, and determining the transverse stacking times N of the block of the feature map according to the types and the number of FPC defect detection; S5, carrying out model training, i. e., sending the marked image data set to the improved MAKS RCNN network model for training; S6, performing performance evaluation on the trained FPC defect detection model; and S7, performing parameter optimization and fine tuning, and performing further optimization on the model in combination with an evaluation result in the step S6. According to the invention, the automatic detection of FPC defects is realized, the defect size can be obtained by segmenting the image defects through the mask, the re-judgment of workers is facilitated, and the detection cost of enterprises is greatly reduced.

Description

technical field [0001] The invention relates to the technical field of defect detection and image segmentation of flexible circuit boards, in particular to an FPC defect detection method based on improved MASKRCNN. Background technique [0002] Flexible printed circuit board is referred to as FPC (Flexible Printed Circuit Board), which has the characteristics of thin thickness, light weight, and free bending and folding. Compared with traditional circuit boards, FPC takes up less space and can greatly reduce package size and weight to meet the requirements of high integration and mobility of electronic products. It can realize three-dimensional space wiring of circuits and enhance product reliability. Reduce assembly costs. Because FPC materials are special, highly integrated, and the process is complex, the manufacturing process is easily affected by factors such as equipment, personnel, and the environment, resulting in defects. [0003] The MASK RCNN network is an impro...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to overcome the deficiencies of the prior art. The present invention provides a FPC defect detection method based on improved MASK RCNN, which realizes automatic detection, classification and segmentation of FPC main defects, saves labor costs of enterprises, and greatly improves Efficiency of defect detection

Images