Eagle eye color cognition antagonism mechanism-imitated unmanned aerial vehicle marine target detection method

Abstract

The invention discloses an eagle eye color cognition antagonism mechanism-imitated unmanned aerial vehicle marine target detection method. The method comprises the following steps of: 1, initializinga channel and setting parameters; 2, setting a mask layer; 3, calculating an eagle-eye-imitating color antagonism mechanism; 4, calculating an eagle-eye-imitated anisotropy Gaussian difference filtering kernel; 5, calculating an eagle-eye-imitated color cognitive antagonism isotropic Gaussian difference filtering kernel; 6, combining the two channels and adjusting the contrast ratio; 7, combiningcontour responses of all channels of the eagle-eye-imitating model; and 8, filtering sea wave interference and performing outputting. The method is oriented to the ship target contour extraction application background in a complex sea environment, solves the target contour detection problem under the conditions of sea surface reflection or scale light sea waves and the like in the complex sea surface environment, suppresses the interference problem of the sea surface scale light sea waves on target contour extraction, and is accurate in contour extraction result. Contour extraction of a marineship target can be realized, and the anti-interference performance and stability of marine target contour extraction are improved.

Description

technical field [0001] The invention relates to an unmanned aerial vehicle sea target detection method imitating an eagle-eye color cognitive antagonism mechanism, and belongs to the field of unmanned aerial vehicle perception and navigation. Background technique [0002] Contour detection is an important basis for technologies such as target recognition, target detection and target tracking. It refers to the process of ignoring the background and internal texture of the target and the influence of noise interference in the digital image containing the target and the background, and adopting certain techniques and methods to realize the process of extracting the target contour. Contour features reflect the shape of objects and can provide important information in vision applications such as object detection, recognition, and tracking. [0003] At present, there are two types of contour detection methods, one is to use the traditional edge detection operator to detect the ta...

AI Technical Summary

Problems solved by technology

[0005] There will be some obvious interference in the detection of the outline of the target on the sea, such as the scale reflection formed by the sunlight on the sea surface, the spray formed by the sea wave movement, and the wake wave interference caused by the ship movement. The reflective part is also detected as a target and retained in the final detection result. Therefore, the traditional contour detection method cannot realize the target contour detection in the complex environment at sea

[0006] In view of the complexity of the sea environment and the many interferences such as scales and waves, the requirements for sea target contour detection are closely related to the uniqueness of the sea environment in the target detection stage and the accuracy, stability, and anti-interference of the target contour detection results. Therefore, the present invention proposes an unmanned aerial vehicle target detection method imitating the eagle-eye color cognitive antagonism mechanism, so as to solve the shortcomings of the existing maritime target contour detection and traditional methods for interference suppression in complex sea environments, and to promote Accurate contour detection of ship targets at sea is successfully realized

Images