Robot target grabbing detection method based on continuous path

Abstract

The invention relates to a robot target grabbing detection method based on a continuous path. The continuous path is obtained on a grabbed objected, geometric center points of overlapped grabbing areas on the grabbed object are connected, and a path set is obtained; a redundancy path is removed from the path set, confidence evaluation is carried out, and the effective continuous path is obtained;on the basis of the YOLO V3 model, a multiscale grabbing detection model of the grabbed object is built; and under a Darknet frame, the multiscale grabbing detection model is trained, an image containing a grabbed object is input to the trained multiscale grabbing detection model, and an output grabbing parameter is obtained. The method has the beneficial effects that firstly, all feasible grabbing area distributions can be described, the grabbing probability of the grabbing area can be more accurately evaluated and predicated, rapid convergence of the grabbing detection model is facilitated,the model training time is shortened, multiple grabbing areas with different positions and different scales can be predicated at the same time, and multiple grabbing chooses are provided for actual grabbing operation.

Description

technical field [0001] The invention belongs to the field of image processing and computer vision, and relates to a robot target grasping detection method based on a continuous path. Background technique [0002] With the development of artificial intelligence technology, robots play an increasingly important role in industrial production and family life. However, existing robots usually only have simple voice interaction and obstacle avoidance capabilities, and cannot carry out higher-level interactions with humans. Robot grasping is an important means to realize human-computer interaction. Through the operating gripper installed at the end of the mechanical arm, the robot can realize the grasping operation of the target, which can be applied to the fields of assembly line sorting, home service and so on. [0003] In order to implement target grasping, it is necessary to define the six-dimensional pose of the manipulating gripper. However, the direct calculation of pose in...

AI Technical Summary

Problems solved by technology

[0004] In terms of grasping area representation, the existing methods usually give several discrete directed rectangular areas as the real value, but this representation method has certain limitations: it only gives a part of the effective grasping representation, and any Deviated areas are considered to be invalid crawl areas, which is not true

Therefore, the existing method of comparing the angle deviation and coverage index between the predicted area and the real area cannot effectively reflect the actual graspable probability of the current area. A method with continuous feature distribution that can reflect all feasible graspable areas of the current object is adopted. It is very important to capture the representation method

Images