基于图像处理技术的扇贝尺寸及位置自动识别  

Automatic measurement of size and location in scallop based on image processing

作　　者：郭常有[1] 曹广斌[1] 韩世成[1] 蒋树义[1] 

机构地区：[1]中国水产科学研究院黑龙江水产研究所,黑龙江哈尔滨150070

出　　处：《大连海洋大学学报》2012年第6期578-582,共5页Journal of Dalian Ocean University

基　　金：国家农业行业专项(201003055)

摘　　要：准确识别扇贝的大小和位置是实现扇贝自动化分级操作的首要条件。本研究中,应用图像处理技术对扇贝进行定位和识别,首先将采集到的扇贝彩色图像转换为灰度图,并通过阈值分割转变为二值图像,再通过滤波减少图像噪声,然后采用改进的OPTA算法和边界追踪算法提取扇贝图像边界,最后通过计算扇贝图像边界各点相对距离的最大值识别其尺寸,通过计算中心点得到位置坐标。实验结果表明,将图像处理技术用于扇贝的自动分类时,能够有效地完成扇贝的定位和尺寸识别,且计算量小,易于实现。It is important to measure size and location of scallop precisely for automated classification for scallops. This paper dealt with a method for automated classification for scallops based on image processing technology. First, the collected original image was converted to grey-scale image, which was then converted to binary image u- sing the threshold segmentation method. Then, filtering method was adopted to reduce image noise. Furthermore, improved OPTA algorithm and tracing algorithm were used to extract the boundary of the scallop image. Finally, the size can be measured by calculation of the maximum distance of the scallop image boundary, and position coordi- nate is obtained by calculating the center point. The experiments showed that the proposed algorithm completed rec- ognition of the positioning and size of the scallops for the automatic scallop classification effectively, with less calcu- lation consumption and easy implement.

关 键 词：图像处理 改进的OPTA算法 中点法 扇贝 

分 类 号：TP241.2[自动化与计算机技术—检测技术与自动化装置]