Fully automatic micro-insect trapping detection method and system thereof

Abstract

The invention provides a fully automatic tiny insect trapping detection method and system thereof. The method includes identifying and marking target insects in the collected images; Separating the target insects and obtaining the first separation and counting results; using the local area approximate estimation algorithm to separate the severely overlapped and occluded target insects in the marked collected images, and obtaining the second separation and counting results; obtaining the results including the first separation and counting Result and the total count result of the second separate count result. The present invention can significantly improve the counting accuracy of target tiny insects, and can separate target insects that are partially overlapped, severely overlapped, or blocked by obstacles, and obtain the counting results that are closest to the facts; at the same time, the image acquisition device of the present application improves Environmental adaptability and image quality of tiny pests.

Description

technical field [0001] The invention relates to the technical field of agricultural detection, in particular to a fully automatic tiny insect trapping detection method and a system thereof. Background technique [0002] With the vigorous development of computer technology and digital image processing technology, the image processing technology of large and medium-sized pests and the automatic counting technology of target pests have been relatively mature, but for tiny insects (such as aphids, planthoppers, tobacco powder, etc.) The identification and counting technology of small agricultural pests such as lice and Liriomyza is still in a relatively lagging stage. On the one hand, due to the large number of tiny insects and their small size, it is difficult for ordinary digital cameras to capture high-quality pictures; It is easy to lose. [0003] Bemisia tabaci is a worldwide vegetable pest, currently distributed in other provinces, municipalities and autonomous regions i...

AI Technical Summary

Problems solved by technology

However, due to the small size of Bemisia tabaci individuals, dozens, hundreds or even thousands of them are often gathered on one leaf when they do damage in the field, which brings trouble to manual investigation work, low investigation efficiency, large investigation errors, and investigation Personnel experience, vision, mood, fatigue, and environmental factors all affect the accuracy and reliability of survey data

Therefore, the accuracy of the sampling survey of Bemisia tabaci in the field has become a bottleneck affecting the ecological research of Bemisia tabaci and the monitoring of diseases and insect pests of Bemisia tabaci

[0004] In the prior art, the patent application whose application publication number is CN 104813993 A discloses a kind of automatic detection device and method for tiny agricultural pests based on machine vision. The detection device is composed of automatic traps for tiny agricultural pests, cameras and computers; In practical application, the process of applying the insecticide, removing the insecticide, and adding the insecticide manually cannot be fully automated, the implementation efficiency of the device is low, and the structure is relatively complicated; at the same time , which uses a digital camera to collect images under natural light sources. For tiny pests with a size of only 1-2 mm, the quality of image collection is not good, and the collection equipment is difficult to adapt to the complex environment in the field

[0005] Further, based on the analysis and processing of the collected images of the prior art, due to the serious deformation of the pests attached to the traps, such as the overlap of the target pests, there will be errors in the counting of pests, so the accuracy of the recognition technology is also limited. needs improvement

Images