Cross-platform vision-guided dispensing guidance method

Abstract

The invention discloses a cross-platform vision-guided glue dispensing guidance method, comprising the following steps: the first step, collecting images; the second step, drawing a characteristic area, and creating a matching positioning template; the third step, teaching glue paths, generating Visual model of dispensing trajectory; step 4, search for matching template; step 5, transfer model to image coordinate system; step 6, transfer image coordinate system to mechanical axis coordinate system; step 7, calculate dispensing trajectory visual model The absolute mechanical axis coordinates of the coordinate points; the eighth step, output the model coordinate points, and the lower computer executes coordinate dispensing. The cross-platform vision-guided dispensing guidance method has the advantage that the dispensing track vision model and its guidance repositioning can be shared between different devices.

Description

technical field [0001] The invention relates to the technical field of vision-guided dispensing, in particular to a cross-platform vision-guided dispensing guiding method. Background technique [0002] In order to realize the precise dispensing operation guided by the industrial camera on the mechanical axis, it is widely used to install the industrial camera on the mechanical axis to realize 2D vision-guided dispensing. The specific steps are: the first step, the industrial camera and the dispensing needle on the mechanical axis Calibration of the positional relationship in the coordinate system; the second step, the teaching of the visual model of the product’s dispensing trajectory; the third step, the guiding repositioning of the product’s dispensing trajectory visual model. [0003] For the second step above, the traditional visual model of dispensing trajectory is composed of "glue road axis coordinate point set" and "image coordinate point in the center of the templat...

AI Technical Summary

Problems solved by technology

[0005] In actual production, a product often uses multiple devices to work in parallel, and the imaging magnifications of cameras between different devices are different. The traditional dispensing trajectory visual model and guidance method are used on different devices, because the imaging magnifications between devices are inconsistent. , the imaging position of the dispensing trajectory vision model in the image will be scaled and transformed with the center of the image as the origin, which will cause the imaging coordinates of the model to translate to the center, while the traditional method does not consider the model translation caused by scaling, which will eventually lead to the guided relocation of the glue road The position is wrong, making it impossible to use the same model to guide dispensing between different devices

Images