【摘要】 为了实现贴片机对SOT元件的高精度、高质量的识别定位,提出了一种基于Hu矩和递进Hough变换的SOT元件识别算法。该算法首先对采集图像进行预处理,在二值化图像的基础上提取元件主体轮廓,采用八邻域跟踪算法对轮廓信息进行提取;然后利用元件主体轮廓的面积和Hu矩,对SOT元件进行识别检测;最后在识别的基础上,对其进行递进hough变换的定位检测。实验表明:该方法可提高元件识别率,且在定位时具有耗时短、精度高等优点,能较好满足贴片机高精度高质量的要求。更多还原

【Abstract】 In order to realize the high-precision and high-quality recognition and positioning of SOT components by the placement machine, a SOT component recognition algorithm based on Hu moment and progressive Hough transform is proposed. The algorithm firstly preprocesses the acquired image, extracts the contour of the component body based on the binarized image, and extracts the contour information by using the eight-field tracking algorithm to obtain the center coordinates of the component. The SOT component is then identified for detection using the area of the component body profile and the Hu moment. Finally, on the basis of identification, it performs the location detection of progressive hough transform. Experiments show that the method improves the component recognition rate and has the advantages of short time-consuming and high precision in positioning, which can better meet the requirements of high quality and high precision of the placement machine.更多还原