A method for processing fishery remote sensing test image data

Abstract

This invention relates to the field of image processing technology, and more particularly to a method for processing fishery remote sensing test image data. The method includes the following steps: acquiring a sequence of remote sensing test images of aquaculture areas in net cages of *Mackerel sanguinea*; extracting the aquaculture area regions from the remote sensing test image sequence and performing brightness gradient elimination to obtain a brightness-normalized image; determining the spatial frequency and texture directionality of water surface fluctuations in the brightness-normalized image to obtain the fluctuation characteristic wavelength and texture direction dispersion; dividing the fluctuation characteristic wavelength by the texture direction dispersion to obtain the water surface dissipation offset index; and extracting the water surface dissipation offset index from multiple consecutive frames to calculate the time series variance. This invention, through the analysis of water surface fluctuation and direction dispersion characteristics in remote sensing test images of aquaculture areas, enables a reliable determination of the continuous activity status and activity level of fish schools in net cage aquaculture areas.

Images