机器视觉的背景颜色对水稻种子图像分割的影响

Effects of Background Color in Machine Vision on Image Segmentation of Rice Seed

在机器视觉识别水稻种子的发展过程中,背景颜色对水稻种子的图像分割直接影响着识别分析的准确性。采用4种不同的背景颜色进行图像采集,即将黑色、蓝色、绿色和红色涂在种子上,分析其图像识别效果,研究背景颜色对水稻种子图像分割的影响。利用LabVIEW软件实现图像处理,对水稻种子的长度、宽度等简单参数进行测量,根据水稻种子的实际长度和宽度计算不同背景色下的图像识别误差。研究发现,蓝色背景下种子长度和宽度的识别对比度更好,识别准确度更高。

In the development of machine vision for rice seed recognition,the background color directly affects the accuracy of the recognition and analysis of the image segmentation of rice seeds.In this paperfourdifferent background colors are used in image acquisition,that is,black,blue,green and red are painted onthe seeds,the image recognition effect is analyzedand the effect of the background color on the image segmentation of rice seeds is studied.By L.abVIEW software to realize image processing,measure simple parameters such as the length and width of rice seeds,and calculate the image recognition errors under different background colors according to the actual length and width of rice seeds.The research found that therecognition contrast of seed length and width under blue background is betterand the recognition accuracyis higher.