Machine Vision Recognition And Analysis For Aluminum Alloy MIG Welding Pool And Metal Transfer

The purpose of the subject is to acquire the useful quality characteristic information from the weld pool and the metal transfer image, which is gotten by ordinary CCD and high-speed camera, and build a good foundation for the welding quality control based on vision sensing.With a passive vision sensing method, a CCD camera was used to obtain the weld pool image of aluminum alloy pulsed MIG welding. And the high-speed camera was used in the process of aluminum alloy Dual Bypass MIG welding to have the weld pool and droplet image. These two kinds of camera both have a narrowband filter. The digital image-processing technology was used to detect the edge characteristic information from the weld pool image by MATLAB. The results were compared and analyzed in order to provide a choice of methods in practice. The digital image-processing technology was also used to detect the edge characteristic information of detachment and droplet from the metal transfer image in aluminum alloy Dual Bypass MIG welding. The results of every method were compared, and a new method of multi-arithmetic operators was invented and acquired clear edge information. The edge information of droplet was acquired by brightness and subtraction based separation algorithm. This will build a good foundation between bypass current and characteristic information of metal transfer for aluminum alloy Dual Bypass MIG welding. In order to check out the results, manual measurement was taken.In order to study the stability of metal transfer in aluminum alloy Dual Bypass MIG welding, the change of the size of necking which was in different welding procedures was analyzed. The change of droplet geometry and velocity in falling process with different bypass current was calculated and analyzed. These are all can be used to help study the affections between bypass current and stability of system and quality control.