A new method for soybean leaf disease detection based on modified salient regions

Abstract

Soybean is the main food crop and an important economical crop of the world. Proper disease control measures must be undertaken to minimize losses. Techniques of machine vision and image processing were applied mostly to plant protection in recent years. Disease detection and segmentation are very important, but the diseases of soybean are complex in real environment and traditional segmentation methods cannot quickly and accurately obtain segmentation results. This research presented a new method for soybean leaf disease detection based on salient regions. This method used low-level features of luminance and color, combined with multi-scale analysis to determine saliency maps in images, and then K-means algorithm was used. The experimental results show that this method can accurately extract the disease regions from soybean disease leaf images with complex background, and it can provide an excellent foundation for extracting disease feature and identifying the diseases categories.