关键词: 机器视觉, 霍夫变换, 透视变换, 标准模板

Abstract: Themanual visual inspection methodis used to detectthe lack of reinforcement dispensing of mounting devices, which has the problems of high labor cost and low efficiency. A method of missing dispensing detection based on machine vision is proposed.The hue-saturation-value(HSV) threshold segmentation method is used to identify the board image. The minimum envelope rectangle is used to extract the target image. The Hough transform and perspective transform are used to obtain the front-view image of the glue dispensing area. Each glue dispensing area is manually marked and segmented,and the area of each glue dispensing is recognized and saved as a standard template.The image of the product to be inspected is processed by the above method, and compared with the standard template,the quality of the product to be inspected is judged and the location of the missing glue position is identified.The method based on machine vision is used to carry out the actual confirmation test. The results show that the method can accurately and reliably detect and judge the dispensing quality of the product, and identify all dispensing missing locations. The accuracy rate of detecting unqualified products is 100%, and some qualified boards have passed the inspection. The comprehensive accuracy rate of detection is 98.5%. It has good accuracy and reliability, and can meet the detection requirements.

Key words: machine vision, Hough transform, perspective transform, standard template