基于自适应模糊阈值的茶炭疽病斑面积计算

Calculation of Spot area of Tea Anthracnose Based on Adaptive Fuzzy Threshold

茶炭疽病是一种发病比较普遍的茶叶病害,它侵蚀茶树叶部,造成茶叶减产和品质下降。利用计算机图像处理技术对茶炭疽病图像进行分割和分析,可以实现病斑面积的自动识别计算,为病害识别提供新的手段。本文提出一种基于自适应模糊阈值方法计算茶炭疽病病斑面积,通过从不同颜色空间模型中使用模糊阈值方法分离目标叶片和病变区域,采用中值滤波和最大连通域技术消除部分边缘噪声干扰,最后根据像素转换的病斑区域和叶片面积计算相对病斑面积。实验通过对不同等级的病斑图像进行实验,以手动分割结果作为标准计算相对准确率和绝对准确率。结果表明自适应模糊阈值方法能够有效计算茶炭疽病病斑面积,准确率达到90%以上,效果稳定,具有较高的准确度和鲁棒性。

Tea anthracnose is a common disease of tea,which erodes tea leaves and leaves,resulting in reduced yield and quality of tea.The computer image processing technology is used to segment and analyze the image of tea anthracnose,which can realize the automatic recognition and calculation of lesion area and provide a new means for disease identification.In this paper,an adaptive fuzzy threshold method is used to calculate the area of tea anthracnose lesions.By using fuzzy threshold method to separate target leaves and lesions from different color space models,median filtering and maximum connected domain techniques are used to eliminate some edge noise interference.Finally,the relative lesion area is calculated according to the lesion area and the leaf area of the pixel conversion.The experiment carried out experiments on different levels of lesion images,and used manual segmentation results as a standard to calculate relative accuracy and absolute accuracy.The results show that the adaptive fuzzy threshold method can effectively calculate the area of tea anthracnose lesions with an accuracy of more than 90%,stable effect,high accuracy and robustness.