摘要
In the field of plant protection,certain methods for assessing the current pest situation and implementing appropriate protection countermeasures can effectively protect plants while saving manpower and material resources.However,current pest monitoring methods are primarily based on the stage of"seeing,hand checking,disc shooting and net catching",and the level of automation is low.Manual methods are time-consuming,prone to error,and difficult to review.We designed a method based on infrared thermography principle for counting Ricania guttata(Walker),a pest which is harmful to mangrove plants.This method,which is based on thermal infrared images and binarized image segmentation,realizes image processing of surface temperature,effectively distinguishes pests and sticky board,automatically counts the number of pests,and expands the data source based on image processing.Furthermore,this method contributes to the solution of the problem that counting error of insect close to the color of sticky board is greater in image recognition of visible light,when the pest color is close to the stick board.It can facilitate manual investigation of mangrove pests,simply and efficiently count the number of pests on the stick board,and provide data and technical support for pest condition analysis and control.
