Recognition method of rov deformation small target based on convolution kernel screening ssd network

Abstract

The present invention provides a target re-identification method based on densely connected convolutional network hypersphere embedding. Firstly, the densely connected convolutional network DenseNet extracts the underwater deformation target features in the video sequence, which greatly reduces the disappearance of gradients, strengthens feature propagation, and supports The process of feature reuse and parameter learning, and then from the perspective of fine-grained classification, from the local integration to the global, using the group average pooling idea to refine and extract the features of underwater deformation targets at all levels, to obtain more accurate underwater deformation target feature expression capabilities, and to The hypersphere loss, that is, the angular triple loss, focuses on the inter-class differences of underwater deformation individual targets, distinguishes intra-class differences, avoids directly measuring the Euclidean distance between the encoding features of underwater deformation individual targets, and constructs a complete underwater vision system with multi-point deployment. A continuous underwater deformation individual target re-identification model. The present invention finally completes the close supervision and process tracking of the underwater deformation target individual in the short-distance multi-field observation.

Description

technical field [0001] The invention relates to a detection method based on small deformation targets, and belongs to the technical fields of intelligent information processing, target detection and underwater robots. Background technique [0002] Underwater small-deformation target detection is an indispensable link in most vision systems. In specific scene applications (such as video surveillance and other fields), automatic, fast, and high-robust target tracking has attracted attention. It has broad application prospects in monitoring, traffic detection, intelligent robots, and submarine target detection and tracking. In addition, based on the strategic significance of the ocean, the ocean must be rationally developed, researched and utilized. [0003] Underwater robot (Remote Operated Vehicle, ROV) can replace human beings to work in complex and dangerous underwater environments due to its flexible maneuverability and strong autonomy, and has been applied to various mar...

AI Technical Summary

Problems solved by technology

To this end, we have introduced an underwater robot control platform based on Raspberry Pi and flight control. For the current deep learning model, the deeper the model contains more parameters, which brings a significant increase in the amount of calculation. The large amount of calculation demand leads to deep learning. It cannot run in embedded devices in real time, and the increase in power consumption is also difficult to solve

Images