A segmentation method for greenhouse vegetable foliar disease spots images using color information and region growing

- A comprehensive color feature and its detection method are proposed.
- The method can segment disease spots images captured under real field conditions.
- The method adopts a region growing method based on CCF map to obtain the disease spots segmentation.
- The method guarantees an accurate input to CNN based disease identification.

This paper presents a novel image processing method using color information and region growing for segmenting greenhouse vegetable foliar disease spots images captured under real field conditions. Disease images captured under real field conditions are suffering from uneven illumination and complicated background, which is a big challenge to achieve robust disease spots segmentation. A disease spots segmentation method consisting of two pipelined procedures is proposed in this paper. Firstly a comprehensive color feature and its detection method are presented. The comprehensive color feature (CCF) consists of three color components, Excess Red Index (ExR), H component of HSV color space and b∗ component of L∗a∗b∗ color space, which implements powerful discrimination of disease spots and clutter background. Then an interactive region growing method based on the CCF map is used to achieve disease spots segmentation from clutter background. To evaluate the robustness and accuracy, the proposed segmentation method is assessed by cucumber downy mildew images. Results show that the proposed method can achieve accurate and robust segmentation under real field conditions.