Method for grabbing object by mechanical hand based on image information

Abstract

The invention relates to a method for grabbing an object by a mechanical hand based on image information. The outline of the object is extracted from the image information, and a containing grabbing manner which can ensure that the object can not escape from the fingers of the mechanical hand is calculated out according to the outline information. The method is characterized in that the direct calculation of the accurate position of the grabbing point is avoided by calculating a containing grabbing area, the object can be ensured not to escape from the mechanical hand and can be finally grabbed firmly. The method comprises the following steps of: firstly, extracting the edge outline of the object from the image information by using an image processing module; secondly, fitting out the minimum convex polygon surrounding the object outline by using a convex hull calculation module; thirdly, calculating out the minimum width of the minimum convex polygon by using a minimum width calculation module; and finally, calculating the relative position relation between the mechanical hand and the object according to a containing grabbing area calculation module when the mechanical hand grabs the object, and guiding the mechanical hand to grab the object.

Description

technical field [0001] The technical field of robots of the present invention relates to a method for grabbing an object by a manipulator based on image information. Background technique [0002] In industrial production activities, the use of manipulators to grab workpieces placed on workbenches or conveyor belts is the basic process for completing sorting, handling or assembly tasks. For example, in an automobile engine production line, it is necessary to use a manipulator to pick out connecting rods from a pile of messy workpieces and assemble them into the engine. On the bolt processing line, nuts that do not meet specifications need to be sorted from the conveyor belt. On the toy production line, toys of different shapes need to be moved into different packing boxes. In the field of life, there are a large number of practical service robots. These service robots need to intelligently grasp objects placed in different positions according to the user's instructions, su...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the high complexity of the grasping algorithm based on model reconstruction, and in the process of use, due to the inaccurate position of the fingers of the manipulator, the object is separated from the manipulator and the grasping is unstable; and, the grasping method based on offline learning requires Training in advance to grasp the mapping function, the real-time performance of the algorithm is not good, etc., the purpose of the present invention is to provide a method for manipulators to grasp objects based on image information

Images