基于随机森林算法的黄瓜种子腔图像分割方法

发布时间：2018-08-26 11:36

【摘要】：针对黄瓜表型测量中图像识别问题,为解决黄瓜种子腔与果肉图像灰度差别不大情况下的分割难题,提出了基于随机森林算法(Random Forest,RF)的黄瓜种子腔图像分割方法。首先,通过颜色空间变换,提取样本在RGB、HSV、YCb Cr模型下的9个颜色分量;接着,基于灰度共生矩阵提取样本的能量、熵、对比度、相关性的均值与标准差等8个纹理特征。结合纹理与颜色特征,运用随机森林算法构建像素分类器,实现了种子腔的粗分割。为了提高分割质量,对粗分割的图像进行形态学处理得到最终分割图像。最后,与K-均值聚类(Kmeans)算法、支持向量机(Support Vector Machine,SVM)算法做对比。实验表明:随机森林分割算法正确识别率高达95%,错误识别率在10%之内,处理时间1.6 s左右,分割质量上优于其它两种算法。
[Abstract]:Aiming at the problem of image recognition in cucumber phenotypic measurement, a method of cucumber seed cavity image segmentation based on stochastic forest algorithm (Random Forest,RF) was proposed in order to solve the problem of segmentation of cucumber seed cavity image with little difference in gray level between cucumber seed cavity and pulp image. Firstly, the 9 color components of the sample under RGB,HSV,YCb Cr model are extracted by color space transformation, and then eight texture features, such as energy, entropy, contrast, correlation mean and standard deviation, are extracted based on gray level co-occurrence matrix. Based on texture and color features, a pixel classifier is constructed by using random forest algorithm to realize rough segmentation of seed cavity. In order to improve the segmentation quality, the coarse segmentation image is processed by morphological processing to get the final segmentation image. Finally, compared with K-means clustering (Kmeans) algorithm and support vector machine (Support Vector Machine,SVM) algorithm. The experimental results show that the correct recognition rate of the stochastic forest segmentation algorithm is up to 95%, the error recognition rate is within 10%, the processing time is about 1.6 seconds, and the segmentation quality is superior to the other two algorithms.
【作者单位】：上海交通大学机械与动力工程学院;上海交通大学农业与生物学院;
【基金】：“十二五”国家科技支撑计划项目(2014BAD08B01) 上海交通大学“Agri+X”基金项目(Agri-X2015002)
【分类号】：TP391.41;S642.2

【相似文献】