A pipeline object location method based on fusion of vision and code disc data

Abstract

The invention discloses a method for positioning a product on an assembly line based on vision and code disc data combined positioning. The method comprises the following steps that for the object on the assembly line in a predetermined space, three cameras are transversely arranged on a support stretching across a conveying belt of the assembly line and acquire images according to an external trigger signal, three images are used as a trinocular image, and the triggering time of the external trigger signal is used as the timestamp; and the images acquired by the left auxiliary camera and the right auxiliary camera are projected onto the image acquired by the main camera, then position coordinates, in an image coordinate system, of all corner points of the object are calculated according to the image obtained after projection and converted into a world coordinate system, a code disc is fixedly arranged on the conveying belt, the conveying belt displacement value obtained according to current data of the code disc is read, the movement vectors of all the corner points from the timestamp moment to the current moment are calculated according to the conveying belt displacement value, world coordinates of all the corner points at the timestamp moment and the corresponding movement vectors are added, and thus world coordinates of all the corner points at anytime are obtained.

Description

technical field [0001] The invention belongs to the technical field of manipulator operation, and in particular relates to an assembly line object positioning method based on fusion positioning of vision and code disc data. Background technique [0002] At present, assembly line operation is playing an increasingly important role in industrial production, and optimizing the assembly line is directly related to the quality and efficiency of products. Under the background that manual sorting cannot meet the needs of large-scale mechanized production, in order to improve production technology and efficiency, an automatic sorting technology is urgently needed to replace traditional manual operations. Among them, the core problem of automatic sorting technology is object positioning. [0003] When using manipulators to grab and assemble parts on the assembly line, visual recognition software is often used to identify the parts and judge their orientation and pose at the same tim...

AI Technical Summary

Problems solved by technology

[0004] Patent CN201210111296.5 proposes a method for positioning objects in a predetermined space by arranging multiple cameras in a predetermined space, but this method needs to use complex algorithms to determine the optimal camera, and only targets objects within a predetermined space

Patent 201420703458.9 adopts an object positioning and grasping platform based on photoelectric sensors. This platform can only complete the grasping of quantitative height objects, which is not universal, and the actual factory operation is expensive.

Moreover, the above two methods are prone to the defect of blind spots caused by the occlusion of the robotic arm when the object is grasped by the robotic arm, and the reliability is not high.

Images