An automatic monitoring method and system for illegal sand mining ships based on underwater acoustic signals

Abstract

The invention discloses an automatic monitoring method and system for an illegal sand mining ship based on an underwater acoustic signal. The method uses a hydrophone to collect underwater acoustic data, and transmits the data to the edge computing module through the signal line of the hydrophone. The underwater acoustic signal processing module amplifies and filters the signal collected by the hydrophone, removes the underwater background noise, amplifies the underwater target voiceprint signal, and sends the processed signal to the target feature judgment module. The judging module extracts the acoustic features of the signal through the signal transformation algorithm, and matches them with the features in the pre-established sand dredger voiceprint feature library. When the matching degree reaches the set threshold, a reminder signal is sent.

Description

technical field [0001] The invention relates to the technical field of automatic monitoring of ships, in particular to an automatic monitoring method and system for illegal sand mining ships based on underwater acoustic signals. Background technique [0002] Target recognition based on acoustic features is one of the important research directions in the field of artificial intelligence at home and abroad in the past two decades. At present, human voice speech recognition technology and animal voice recognition technology in this field have matured and been widely used. In terms of ship voiceprint feature determination, a mature method has been formed based on ship engines and ship types. In the process of ship design and construction, the method of testing water pools is used for voiceprint sampling, and a voiceprint library is established, followed by ship feature determination applications. method. On the other hand, the acoustic feature extraction and recognition of sur...

AI Technical Summary

Problems solved by technology

[0003] The existing underwater listening technology and VTS monitoring technology mainly have the following problems: First, the existing underwater listening technology cannot distinguish and filter the background sound of the sand mining ship and the inland river environment, and cannot effectively establish a sand mining environment with a pure background. Acoustic feature acquisition environment of ships is difficult to be practical in practical applications

The second is to establish a voiceprint library based on the characteristics of ship engines. It is necessary to consider the engine characteristics of sand mining ships with different power and different installation positions. Most of the illegal sand mining ships are illegally refitted, and rely on different ship types, so it is impossible to establish coverage. Voiceprint library for all boat types

The third is that VTS monitoring technology mainly combines radar echoes and AIS non-registered information to determine the information of suspected sand mining ships. However, at night, ships have behaviors such as anchoring, resting, and slowing down, which are confused with sand mining behavior.

At the same time, the operating rate of AIS equipment in inland rivers is generally less than 20%, so the number of ships that AIS can participate in judging is not enough to support the application

Images