【摘要】 工业测量中经常需要从获取的光斑图像中提取光斑区域,计算光斑中心的坐标。针对光斑图像具有背景区域较暗且面积大,目标(光斑)区域较亮且面积小的特点,提出了一种适用于光斑图像的阈值分割方法,计算单个光斑中心坐标的重心法,以及采用连通区域标记和区域大小排序计算多个光斑中心坐标的方法。与其他阈值分割方法比较后的实验结果表明,所提出的方法可更好地分割光斑图像,与期望的人工选取的阈值最接近,计算得到的光斑中心坐标准确,且运行时间较短。更多还原

【Abstract】 It is often asked to extract the regions of light spot from light spot-image,and calculate their center coordinates in industry measurement.The light spot-images are characterized by darkand large background,bright and small target(spot).Proposes a thresholding segmentation algorithm for light spot-image,a method to calculate a single spot coordinates using center of gravity,and presents a method to calculate the coordinates of multiple light spots using connected component labeling and sorting based on size of regions.Experimental result shows that the proposed method can obtain better performance of segmenting regions of light spot from light spot-image than other thresholding segmentation methods,and its result is closest to that of thresholding segmentation of artificial selection.The method can accurately calculate the coordinates of light spots and it has lower computation time.更多还原