Detection of crop pests under different weather conditions
In order to improve the detection of crop pests in the natural environment and realize the significance of comprehen-sive management,it is currently difficult for traditional machine vision technology to detect them effectively.In this paper,the super-G absolute value method is used to convert the color image sequence into grayscale to realize the separation of pests from the background.The obtained video image sequence is used with the weather type recognition algorithm to design an adaptive thresh-old coefficient δ for pest detection in different weather environments,and it is fused with the maximum inter-class variance method to realize pest detection.The experimental results show that the obtained pest detection algorithm can effectively eliminate noise and has a good effect of extracting pest information.This method can provide new ideas for pest detection and analysis in intelligent video surveillance.
pests detectionsuper G absolute valueweather type recognition algorithmadaptive threshold coefficient δ
万方数据
