Method and system for extracting disease region of interest based on block marking

Abstract

The invention provides a method and a system for extracting a disease region-of-interest based on a block tag. The method comprises the following steps: dividing an image into a plurality of blocks by extracting edge features in the gray image of an original image; grouping the blocks which are selected from the plurality of blocks and are in accordance with color features of a disease region-of-interest into the disease region-of-interest in a color image of the original image; acquiring a binary image distinguishing a disease region-of-interest from a non disease region-of-interest based on the original image; dividing a non disease region-of-interest in the binary image into a plurality of regions based on distance transform and watershed transform; and grouping the regions which are selected from the plurality of regions and are in accordance with color features of the disease region-of-interest into the disease region-of-interest in the color image of the original image. By the adoption of the method and the system, accurate extraction of the disease region-of-interest can be realized in a field environment.

Description

technical field [0001] The invention relates to agricultural engineering, in particular to a method and system for extracting disease regions of interest based on block marks. Background technique [0002] With the improvement of computer software and hardware performance and the advancement of machine vision technology, the use of machine vision to realize automatic disease identification has attracted extensive attention and has been widely used in the field of plant protection. The research on machine vision technology in agriculture started relatively late, among which the efficient and high-precision segmentation algorithm is a hot and difficult issue for automatic recognition. [0003] The disease area of ​​interest includes normal leaf area and diseased area, which is an important basis for measuring the degree of disease incidence and key information for precise spraying. However, in the field environment, due to the complex environment and uncertain lighting condit...

AI Technical Summary

Problems solved by technology

However, in the field environment, due to the complex environment and the uncertainty of lighting conditions, the gray scale range of the image often overlaps, which will bring some problems to the commonly used threshold method and other fixed standard segmentation methods based only on image color characteristics.

For example, since it is difficult for this type of method to distinguish whether the color difference between the two areas is caused by the surrounding light or the color difference itself, there may be mis-segmentation, so that the region of interest of the disease cannot be accurately extracted

Images