基于视觉的太阳能电池片位置检测

Vision-based Detection of Position for Solar Cells

针对太阳能电池片位置检测算法实时性差和检测精度低等问题,采用归类法检测电池片偏转角度,亚像素二次遍历法对电池片定位。首先,对电池片边缘直线进行归类并从中获取一条直线,对该直线进行亚像素细分,获得该直线极坐标系下的参数,并计算电池片偏转角度;接着,对电池片图像边缘进行粗霍夫变换,求得对称中心线,然后对其进行亚像素细分和第二次遍历,获得中心线极坐标系下的参数并求得电池片中心点。通过实验比较发现,相比较传统算法,该算法在提高检测精度的同时也缩短了检测时间,解决了传统算法实时性差和检测精度低的问题。

Due to the problem that poor real-time and low detection accuracy of the traditional algorithm for the detection of position for the solar cells,using the method of classification and sub-pixel secondary traversal method to detect cell deflection angle and positioning of the battery in this paper.Firstly,categorizing the edge lines and taking a straight line from them,subdividing the line by sub-pixel,obtaining parameters in the linear polar coordinate system and calculating the deflection angle of the cell;then using the Hough transform to process the image,calculating the symmetrical centre line,and then sub-pixel subdivision and second traversal for it,getting the parameters of the centre line polar coordinate system and the centre point of the battery chip.Experiments proved that the algorithm can accurately detect the deflection angle of the battery and accurately position the cell,which solves the problem of poor real-time and low precision of traditional algorithms.