Metric space and image feature space fusing mixed vision servo method

Abstract

The invention relates to a metric space and image feature space fusing visual servo method. The method comprises the steps of establishing an image jacobian matrix; establishing a depth jacobian matrix; fusing the image jacobian matrix and the depth jacobian matrix to obtain a mixed vision servo jacobian matrix; and obtaining the motion control capacity of a robot on the basis of the mixed visionservo jacobian matrix through a control algorithm so as to drive the robot to move to a target location to complete accurate positioning. Through the metric space and image feature space fusing visualservo method provided by the invention, the problems of the singular value, the local minimum of a traditional jacobian matrix and the like are avoided; the complementary image feature space and metric space are fused; and under the condition of information non-redundancy, the deficiency of vision information is made up. A system acquires data in real time and can quickly complete a vision servotask through a simple control algorithm.

Description

technical field [0001] The invention relates to the technical field of visual servoing, in particular to a hybrid visual servoing method for fusion of distance space and image feature space. Background technique [0002] Visual servo technology is one of the research hotspots in the field of intelligent robot development and has been widely used in industry. In order to enable the robot to complete more complex tasks, the robot needs a control system with good performance and the ability to perceive changes in complex environments. [0003] In the traditional control method, the control accuracy of position-based visual servoing (PBVS) depends on the calibration of the model and the calibration results of the camera, which is less robust and requires frequent calibration. Image-based visual servoing (IBVS) is more robust to model parameters and image noise than position-based visual servoing. In the control method based on the development of IBVS, most algorithms rely on t...

AI Technical Summary

Problems solved by technology

In the control method based on the development of IBVS, most algorithms rely on the two-dimensional image feature information obtained by the visual sensor, and then use the obtained image information to complete the servo control. For the three-dimensional space servo task, there is a lack of information about the robot's terminal posture. amount, resulting in insufficient environmental awareness, and the servo effect cannot be further improved

Images