Research On Detection Algorithm Of PCB Board Directional Components

The correctness of the PCB board component plug-in determines the quality and production cost of the circuit board,especially the reverse connection of components with polarity and directivity will directly damage the circuit board and cause loss to production.The defect detection technology of PCB components based on Automatic Optical Inspection(AOI)is gradually replacing manual labor due to its advantages of stable performance and high efficiency compared with manual visual inspection.Therefore,it is of great significance to research and apply to defect detection with directional components on PCB boards.This paper focuses on the defect detection of directional components on PCB boards.It mainly researches the polarity detection of electrolytic capacitors,the direction detection of various sockets and the polarity detection algorithms of diodes.Defect detection of directional components.The main research contents of this article are as follows:(1)First,the initial setting of image brightness is studied.The image brightness parameter index is obtained by calculating the mean and variance of the image pixel deviation value,and the image is calibrated according to the brightness parameter.The lens aperture and camera exposure are adjusted according to the standard value,thereby reducing the algorithm detection performance due to different ambient brightness.influences.(2)The stitching algorithm of PCB board images is studied.When the camera collects images,due to the different angles of the components at different positions,the higher components have occlusions and shadows.To avoid this phenomenon,this article uses stitching to obtain PCB board images.And using the template matching stitching algorithm,when there are higher components in the overlap area,the stitching position will be shifted.Aiming at this problem,this paper proposes a difference matching algorithm,and uses the template matching algorithm to perform secondary matching,so as to find the image overlap area and realize the circuit board splicing.The experimental results prove that the method can reduce the occurrence of splicing position shift.(3)An electrolytic capacitor polarity detection algorithm based on contour features and window smoothness is proposed.Because the electrolytic capacitors on the PCB board are skewed and interfere with white printed text,which affects the positioning and polarity judgment of the inner circle,the traditional algorithm is difficult to adapt to the detection of complex scenes.Therefore,this paper uses approximately circular shape descriptors and polylines.The angular continuity feature is used as the feature information for screening the inner contour of the electrolytic capacitor,and the latent circle in the image is extracted by the generalized Hough transform circle detection algorithm,and then the weight of the inner circle of the electrolytic capacitor is determined by combining the contour area information to determine the proportion.The sliding window is used to calculate the smoothness of the electrolytic capacitor ring unrolled image,and the interference of white characters in non-polar areas is eliminated to realize the electrolytic capacitor polarity detection.Experimental results show that the algorithm in this paper is suitable for a variety of complex scenes.Compared with the traditional Hough transform algorithm and SURF algorithm,the positioning is more accurate,and the accuracy of polarity detection can reach 96.72%.(4)The direction detection algorithms of three different types of sockets are studied.For the detection of color sockets,the socket positioning is achieved through the difference calculation of a single channel image,the minimum area is used to correct the angle,and the socket direction is judged based on the projected histogram.Aiming at the detection of white sockets,this paper retains the socket feature information by means of secondary threshold extraction and multiple segmentation,uses rectangular detection to determine its main body area,and implements socket direction detection based on dark channel images.For the detection of groove sockets,this paper uses high-contrast retention to enhance the image edge information,and then extracts the directional features by calculating the contour similarity to determine the socket orientation.The effectiveness of the three algorithms was verified through experiments,and their detection accuracy rates were 99.1%,97.35%,and 95.23%,respectively.(5)The polarity detection algorithm of the diode is researched,and a diode polarity detection based on improved image matching and gradient edge features is proposed.The coarse matching is performed by calculating the similarity of the color histograms of the two images to obtain a matching frame set,which is then combined with the SSIM structural similarity to extract the best matching result.Then,the edge image of the polar region is obtained by using the gradient edge feature,and the maximum continuous range of pixels in all columns of the image is calculated to determine the polarity direction.Experiments show that the algorithm in this paper is better than the comparison algorithm,and it is robust to diode detection in the presence of printed text and its own reflective interference.The accuracy of polarity detection is 96.12%.This paper proposes corresponding detection algorithms for directional components on PCB boards,and verifies the effectiveness of the proposed algorithm,which can be applied to more complex detection environments,and the detection accuracy rate meets the requirements of PCB board component defect detection.