Bird and frog intelligent monitoring method based on deep learning

Abstract

The invention discloses a bird and frog intelligent monitoring method based on deep learning, which is used for remote wireless ecological monitoring. The method comprises: arranging an outdoor soundacquisition device in a monitoring area, wherein the outdoor sound acquisition device is triggered by a sound triggering module to acquire a bird sound and a frog sound; allowing a wireless communication module to wirelessly transmit an audio file of the bird sound the frog sound to a computer server; allowing the computer server to convert the audio file of the bird sound and the frog sound intoa visual sonogram and performing pre-processing; allowing the computer server to perform image recognition and statistical analysis on the sonogram; and generating a daily monitoring form. The bird and frog intelligent monitoring method compensates for the problems that in manual ecological monitoring, the human and material cost is high, the monitoring time and frequency are limited, the data processing efficiency is low, and requirements of species for professional skills of a recognition person are high, and can be used for long-term ecological monitoring of natural protection sites.

Description

technical field [0001] The invention relates to the technical field of sound intelligent recognition and monitoring, in particular to a method for collecting and intelligently recognizing bird and frog sounds based on deep learning. Background technique [0002] Biodiversity survey and monitoring is one of the daily tasks in nature reserves, forest parks and other nature reserves. However, due to the accessibility of monitoring sites and the tendency of animals to avoid humans, biodiversity monitoring often consumes a lot of manpower and material resources, and may not be able to obtain true and effective investigation and monitoring data. In order to fully grasp the dynamic changes of biodiversity, infrared trigger camera trap technology has been widely used in the continuous collection of wild images of mammals and some birds, but upper layer birds, small or fast-moving birds in forests, and Frogs, being cold-blooded animals, cannot be recorded by infrared cameras, and so...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose a method for intelligent monitoring of frogs and birds based on deep learning to solve: (1) the problem that monitoring time and monitoring frequency are extremely limited in manual monitoring technology; The problem that most birds and frogs cannot be recorded; (3) The problem of high professional skills and low recognition efficiency when manually identifying species

Images