论文部分内容阅读
Most existing image-based methods for crop disease diagnosis usually have high requirement of the input images, including simple background, sufficient depth of field, uniform illumination, etc.And before extracting the lesion areas, these methods always need to remove the complex background, which is often hard to obtain the desired results.Besides, when the lesion areas are small, the captured micro images are always have the low depth of field (DOF), which cannot be dealt with effectively by these methods to extract the accurate lesion areas, causing difficulties to the subsequent disease diagnosis.In order to solve these problems, the paper proposed a method of lesion image acquisition based on image saliency detection and image fusion.Firstly, by integrating the structural features and color features extraction and feature space quantization, the saliency region of crop disease images was detected.The lesion areas can be extracted without the preprocessing of removing the complex background.Meanwhile, to deal with the crop diseases images with the low depth of filed, the focused area was extracted and then based on the multi feature matching, the lesion areas extracted from the image sequence are fused so that the image with the super depth of field which covers all lesion areas can be obtained finally.The images of various diseases of cucumber and rice were used in the experiments.The experimental results showed that our method was better than the threshold method and the graph cutting method in image segmentation of the crop disease images, especially with the low depth of field.Our method can effectively extract the lesion areas from the crop disease images with the low depth of field, which lays a good foundation for the follow-up feature extraction and disease diagnosis.