A color image preprocessing method for rice disease and insect pest identification and counting

Abstract

The present invention is a color image preprocessing method for identifying and counting rice diseases and pests. The same target is simultaneously photographed by a photographing device to obtain a depth image and a color image, and gray value enhancement processing and area segmentation are performed on the depth image to obtain binary values. Image and its connected region Q k (k=1,2...); extract the three components of R, G, and B in the color image, segment the color image, and obtain the connected region R with the characteristics of the color components i (i=1,2...); Calculate the distance between the centroids of the two connected regions, and use this as the preliminary screening condition for the connected regions corresponding to the two images. k and R i Do differential processing, when the differential result is less than the set threshold, Q k and R i The corresponding area is completely matched, that is, the target pest area. The invention realizes the removal of the negative influence of the accumulation of sundries and dust on the image preprocessing process, and facilitates the subsequent identification and counting of pests in the image.

Description

technical field [0001] The invention relates to a method for classifying and counting agricultural rice diseases and insect pests, in particular to a pest identification algorithm and pest counting. Background technique [0002] Rice has an output of more than 500 million tons in the world, with a cultivation area of ​​about 150 million hm2, and is one of the most important food crops. China's rice sown area ranks second in the world. The annual rice output in China reaches 200 billion kilograms, accounting for more than half of the world's output. The average yield per unit is 1.6 times the world level, making a great contribution to the world's rice production. Rice, as the main food crop in my country, occupies an extremely important position in food security. The annual rice planting area is about 30 million hectares, accounting for 30% of the national grain planting area and 20% of the world's rice planting area; the total rice output is nearly 200 million tons, accoun...

AI Technical Summary

Problems solved by technology

[0005] Based on the above problems, in order to solve the problems faced by image recognition, the present invention proposes a color image preprocessing method for identifying and counting rice pests and diseases, through which the algorithm can effectively remove the impact of debris accumulation on image preprocessing, and then Can improve the success rate of pest counting and identification;

Images