Abstract:

In order to identify and monitor the hidden insects in the Erhai wetland, this paper, with the help of the Internet designed
an intelligent insect recognition and real- time monitoring system based on the multifeature insect recognition algorithm. The
information of insect characteristics was collected by infrared radiation sensors, ultrasonic sensors, sound sensors, color sensors and
image sensors. Under the control of the single chip computer, the characteristic data was transmitted to the cloud server through the
network. Then the intelligent recognition was realized by analyzing the characteristics such as life, sound and color. Performance tests
show that: compared with image edge detection and correlation feedback method, it has some advantages in time and space complexity,
and the method is feasible.

Key words: internet of things, wetland insects, intelligent recognition, real-time monitoring, multi-feature extraction