Visual sensor based remote drosophila monitoring system and monitoring method thereof

Abstract

The invention discloses a visual sensor based remote drosophila monitoring system. The monitoring system comprises a solar power supply device, a drosophila trapping device and a control monitoring device; the solar power supply device is composed of a solar panel, a support frame and a solar battery; the drosophila trapping device comprises a top cap, a transparent communicating piece, a trapping bottle, a protective shell and a camera, and the top cap is connected with the trapping bottle through the communicating piece; the communicating piece is mounted in the protective shell, the camera is mounted on the inner wall of the protective shell, and a camera head of the camera faces the communicating piece; the control monitoring device comprises a Fit-pc microcomputer, a 3G (third-generation) communication module and a remote server, the remote server is connected with the Fit-pc microcomputer through the 3G communication module, and self-developed drosophila monitoring and counting software for realizing automatic monitoring and counting is mounted and runs in the Fit-pc microcomputer. The invention further discloses a monitoring method of the monitoring system and a drosophila monitoring and counting method. The monitoring system is simple and reasonable in structure, data are processed and analyzed in real time remotely, and intelligentization of drosophila monitoring is improved.

Description

technical field [0001] The invention relates to the field of pest monitoring, in particular to a visual sensor-based fruit fly remote monitoring system and a monitoring method thereof. Background technique [0002] Fruit fly monitoring is the basic work of plant protection and even agricultural production. At present, with regard to the field monitoring of fruit flies, the number of fruit flies is detected manually by setting up observation points for adults, trapping in the field, and collecting rotted and fallen off fruits, and statistics are made on their growth patterns and distribution. This method is prone to errors, and is limited by the intensity of manpower use. It cannot conduct on-site, real-time, and large-scale rapid monitoring of the occurrence of pests in orchards. As a result, the measurement results cannot reflect the real-time and dynamic changes of large-scale orchard pests. [0003] In the monitoring network, the use of video sensors to obtain dynamic im...

AI Technical Summary

Problems solved by technology

Although computer vision technology is becoming more and more mature, it has not been widely used in agricultural informatization for the following two reasons: on the one hand, it is difficult to obtain high-quality images and videos in the process of agricultural production. The images are all obtained under the ideal environment based on the experiment, which cannot be compared with the field environment; on the other hand, because the characteristics of the agricultural image targets that need to be processed are not obvious, usually only a series of complex image processing algorithms can be used for specific targets. , so it is difficult to achieve real-time processing and real-time feedback information

[0004] However, in the prior art, there are also problems of repeated counting and missing counting in the monitoring and counting method.

For example, the invention patent application with the Chinese patent application number 201310052396.X discloses "a new type of remote real-time intelligent monitoring system for pests and its system monitoring method". The generation of high-voltage killing causes voltage changes in the circuit, and current changes occur, and the current changes are counted once. However, if more than one pest touches the grid structure in the pest control device at the same time, it is still only counted once, resulting in a missing count; Chinese patent application The invention patent No. 200920190971.1 discloses "an automatic counting system for wick trapping", which uses infrared detectors to count pests. Because insects have the behavior characteristics of crawling back and forth, it can cause repeated counting by infrared detection, and the accuracy of data statistics Low

Images