基于统计模式识别的植物病害彩色图像分割方法

A Color Image Segmentation Method for Plant Disease Diagnose Based on Statistical Pattern Recognition

根据植物病害彩色图像的特点 ,提出了用颜色空间作为特征空间 ,利用统计模式识别的监督分类方法。采用基于Fisher准则的线性判别函数来对彩色图像进行真彩色二值化分割。结果表明:该方法具有快速、有效、准确等优点 ,对彩色植物病害图像的分割处理十分有效。

A supervision and sorting technique by means of statistical pattern recognition and using color as characteristic space was put forward according to the properties of plant disease image.The real color binaryzation segmentation for the color image was carried out according to Fisher criterion.The experimental result showed that the method of segmentation had a high speed with high accuraˉcy.It was very effective in segmenting and processing plant disease image.