Deep convolutional neural network-based submerged oil sonar detection image recognition method

Abstract

The invention relates to the technical field of bottom oil detection, in particular to a deep convolutional neural network-based bottom oil sonar detection image recognition method. The method comprises a sonar detection image preprocessing module and a deep convolutional neural network bottom oil target identification module. According to the method, a deep convolutional neural network algorithmis introduced into bottom oil detection and identification; the key problems of automatic sinking-to-bottom oil target identification, automatic leakage position positioning, automatic sinking-to-bottom oil pollution area estimation and the like are solved. The method provides a basis for emergency decision and disposal of marine oil spill accidents, improves the level of a bottom oil detection and identification technology in China, provides technical support for national marine safety and petroleum safety, and has important engineering significance and application value.

Description

Technical field: [0001] The invention relates to the technical field of detection and identification of sunken bottom oil, in particular to a sonar detection image recognition method for sunken bottom oil based on a deep convolutional neural network. Background technique: [0002] Energy is the economic lifeline of a country and an important support for the continuous growth of the national economy. With the development of my country's economy and technology, the country's demand for energy such as oil and gas is also increasing. According to statistics, my country's offshore oil and gas reserves are extremely rich. The geological reserves of petroleum in the South China Sea alone have reached 23-30 billion tons, of which 70% are stored in deep sea areas. With the continuous expansion of the scale of offshore oil development and utilization, oil spill accidents have also increased. The main causes of marine oil spill accidents include: collisions of transport ships, accide...

AI Technical Summary

Problems solved by technology

[0010] The technical problem to be solved by the present invention is that in the process of sonar detection and identification of sinking oil, there are many problems in the detection data such as high noise, difficult to interpret, difficult to distinguish, many manual participation, and low identification efficiency.

Images