Plant protection unmanned aerial vehicle ground monitoring system and monitoring method thereof

Abstract

The invention relates to a plant protection unmanned aerial vehicle ground monitoring system and a monitoring method thereof, and the monitoring system takes an unmanned aerial vehicle as a carrying platform and machine vision as a core, and carries out the statistics of basic seedlings of a farmland. The monitoring system specifically comprises a map acquisition system for shooting a picture setand splicing the picture set; a region picture processing system which is used for carrying out region marking and preprocessing on the spliced plant protection map; a convolutional neural network system which is used for identifying crops and weeds; and a statistical system which is used for counting the basic seedling number of the crops in the area so as to estimate the whole basic seedling number. According to the method, a plurality of areas are randomly selected from hundreds of mu of farmland through a visual platform carried by an unmanned aerial vehicle for identification statistics,and basic seedling information of the whole farmland is estimated. Picture sequences are collected through a machine vision platform carried by an unmanned aerial vehicle and spliced into a panorama of a target farmland, then basic seedlings in the images are recognized through an algorithm, and statistics is completed.

Description

technical field [0001] The invention relates to the field of unmanned aerial vehicles, in particular to a plant protection unmanned aerial vehicle ground monitoring system and a monitoring method thereof. Background technique [0002] At this stage, the statistics of basic seedlings are usually counted manually. With the development of the economy and the progress of science, the traditional statistical methods are time-consuming and labor-intensive. At present, machine vision has achieved remarkable success in many fields such as people flow and traffic statistics. However, its application in basic seedling statistics is still in its infancy. The key to counting basic seedlings is to detect crop plants in pictures or videos. The application and research of machine vision in crop detection is relatively mature. [0003] The traditional measurement method is that researchers or agricultural operators come to the farmland in person, randomly select several areas in hundreds ...

AI Technical Summary

Problems solved by technology

Such a statistical method requires a lot of labor costs, and personnel activities in the field may cause damage to the seedlings

Images