Universal image visual inspection robot

Abstract

The invention discloses a universal image visual inspection robot which comprises the following parts mounted in a robot body: a precision six-axis manipulator, a combined light source, a high-speed camera, a computer and an industrial control system, wherein the computer is connected with and controls the combined light source and the precision six-axis manipulator; the industrial control system controls the operation of the precision six-axis manipulator and opening / closing and selection of the combined light source; the high-speed camera is connected with the computer and instantaneously transmits the image information to the computer. The universal image visual inspection robot disclosed by the invention has the advantages that 1) the production cost is reduced; 2) missing detection is eliminated; 3) the universality is high.

Description

technical field [0001] The invention relates to the technical field of visual inspection, in particular to an all-round image visual inspection robot. Background technique [0002] Existing machine vision inspection of precision parts is mainly divided into two types: one is planar (i.e. 2D) inspection, which uses a camera to fix the plane flow of the product and can only take a few photos of the fixed surface of the product at a time. Comparison, so as to achieve two to three size or appearance detection, the other is the laser scanning method (ie 3D) for the overall modeling and imaging of the measured object, which can detect more sizes, but the detection time will be longer, The modeling time of a medium-sized product in a complex program will be more than 5 minutes, and at the same time, it can only detect the size of the scanned surface and cannot judge the appearance. In reality, there are many sizes and appearances that need to be controlled and controlled in the pr...

AI Technical Summary

Problems solved by technology

[0002] Existing machine vision inspection of precision parts is mainly divided into two types: one is planar (i.e. 2D) inspection, which uses a camera to fix the plane flow of the product and can only take a few photos of the fixed surface of the product at a time. Comparison, so as to achieve two to three size or appearance detection, the other is the laser scanning method (ie 3D) for the overall modeling and imaging of the measured object, which can detect more sizes, but the detection time will be longer, The modeling time of a medium-sized product in a complex program will take more than 5 minutes, and at the same time, it can only detect the size of the scanned surface and cannot judge the appearance

In reality, there are many sizes and appearances that need to be controlled and controlled in the production of precision parts. For an ordinary FPC connector used to connect screens in mobile phones and computers, more than 500 dimensions need to be measured, and 20 appearance points. Existing machine vision inspection equipment is far from satisfying

Therefore, it has to rely heavily on manual inspection, which is inefficient and expensive

Images