Method for labeling sea cucumber target detection result by using rotatable boundary frame

Abstract

The invention belongs to the field of underwater target detection, and specifically relates to a method for labeling a sea cucumber target detection result by using a rotatable boundary frame. The method comprises the following steps of: performing data expansion on a sea cucumber training data set manufactured by using labeme software; constructing a full convolutional neural network; Carrying out offline training on the constructed full convolutional neural network by utilizing the expanded data set; Inputting the image containing the sea cucumbers into the trained full convolutional neuralnetwork to obtain a segmented image; Performing corrosion and filtering operations on the segmented image to obtain a post-processing segmented image; And searching a maximum connected domain on the post-processing segmentation map, namely the detected sea cucumber target. According to the method, the obtained segmented image is corroded, burrs on the periphery of the sea cucumber are removed, itis guaranteed that the minimum circumscribed rectangle is more accurate, the sea cucumber grabbing pose cannot appear outside the sea cucumber, and positioning is more accurate.

Description

technical field [0001] The invention belongs to the field of underwater target detection, and in particular relates to a method for marking sea cucumber target detection results with a rotatable bounding box. Background technique [0002] Sea cucumbers live in seawater below 6 meters. At present, sea cucumber fishing is mainly done by divers, mainly in spring and autumn when the temperature is low, seawater pressure is high and temperature is low. Divers who have been engaged in fishing for a long time There will be occupational diseases such as arthritis, and the artificial fishing efficiency is low and the risk factor is high. With the increasing demand for sea cucumbers, underwater robots are more and more required to automatically catch sea cucumbers, and detecting sea cucumber targets and determining the grasping pose of sea cucumbers are one of the basic tasks for automatic fishing by underwater robots. [0003] At present, the techniques for detecting sea cucumber ta...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the object of the present invention is to provide a method for marking sea cucumber target detection results with a rotatable bounding box, which mainly uses a fully convolutional neural network image The segmentation method is combined with the traditional image processing method to solve the problem that the application scenarios of the existing image segmentation method are limited and the real-time performance is not high. Reflect the problem of the rotation angle of the target, accurately mark the target detection results for the vision-based underwater robot, and determine the grasping position and grasping angle for the two-finger robotic arm installed on the underwater robot

Images