Orchard pest intelligent positioning and identification method suitable for small target sample

Abstract

By analyzing the characteristics of orchard pests and combining deep learning and computer vision technologies, small target samples of the pests in a color image range are positioned and recognized,meanwhile, the growth distribution of the pests is calculated, a certain growth condition is jointly positioned through multiple pests, and factors causing easy growth of pests in the current environment are judged, so that the invention has important significance for promoting development of the fields of agricultural and forestry pest identification and prevention and control.

Description

technical field

[0001] The invention relates to the technical field of intelligent detection and identification of agricultural and forestry pests, in particular to an intelligent positioning and identification method for orchard pests suitable for small target samples. Background technique

[0002] At present, my country's agricultural modernization level has been continuously improved and further developed, and the consumption level of agricultural products has been upgraded rapidly, and my country has become one of the countries with the largest scale and variety of fruit production and planting in the world. However, fruit trees are often infested by insect pests in the process of growth, and the quality and yield of fruit have been affected to varying degrees, which has brought huge economic losses to the fruit industry. Therefore, in order to minimize the losses caused by pests, it is necessary to timely and accurately identify and forecast pests in orchards.

[0003] ...

Images