基于颜色模型和阈值分割的有遮挡的柑橘果实识别算法

李  欣; 王玉德

基于颜色模型和阈值分割的有遮挡的柑橘果实识别算法

引用本文：李欣，王玉德.基于颜色模型和阈值分割的有遮挡的柑橘果实识别算法[J].计算技术与自动化,2022,(2):136-140

摘要点击次数: 308

全文下载次数: 0

作者	单位
李欣，王玉德	(曲阜师范大学网络空间安全学院，山东曲阜 273165)

中文摘要:针对柑橘果实被枝叶遮挡，机器识别难度大的问题，提出了基于YCbCr颜色空间模型和OTSU阈值分割技术对有遮挡的柑橘果实进行识别的方法。应用边缘检测法和凸包算法提取有遮挡果实的轮廓特征，选取凸壳上距离最远的两点，作为果实直径的两端点，定位目标果实。结果表明，算法可以准确分割并识别出枝叶遮挡的柑橘果实，果实平均检出率达到90.48%，有效提高了果园自动化采摘的效率。

中文关键词:遮挡柑橘果实目标定位数字图像处理 YCbCr颜色模型 OTSU阈值分割法

Occluded Citrus Fruit Recognition Algorithm Based on Color Model and Threshold Segmentation

Abstract:In order to reduce the difficulty of machine recognition of citrus fruit obscured by branches or leaves, a recognition approach of obscured citrus fruit was proposed based on the YCbCr color space model and the OTSU threshold segmentation technology. The contour features of the fruit obscured by branches and leaves were extracted by using the edge detection method and convex hull algorithm. Two furthest points on the convex hull were selected as the two ends of the fruit diameter to locate the position of the target fruit. The results showed that the algorithm could accurately segment and identify the images of citrus fruits covered by branches and leaves, and the average detection rate of fruit reached 90.48%, effectively improving the efficiency of automatic picking in orchards.

keywords:occlusion of citrus target target location digital image processing YCbCr color model OTSU threshold segmentation method

查看全文 查看/发表评论 下载pdf阅读器