The Collecting And Character Analysis Of The Welding Pool Image Based On The Doulde Wire Arc Welding Process Of The Aluminum Alloy

A vision detecting system of collecting image of welding pool was designed in viewof the MIG of aluminum alloy in this thesis. This detecting system which was composed ofCCD camera, the compound filter system, a image gathering card and a computer can getthe image not only from the behind of the welding pool, but also from the side of thewelding pool. Through the welding process of the 3A21 aluminum alloy sheet, clearimages could be collected both from the behind of the welding pool and the beside of thewelding pool.The image processing algorithm in view of the image which was gathered from theprocessing of the double wire automatic gas shielded arc welding was designed anddeveloped in this thesis. The processed welding pool images disposed by this algorithmwere not only slick and consecutive but also tranquil. The accurate shape characteristic canbe got from the disposed images. After the thorough gray-scale analysis and geometrymorphological analysis for the images of the welding pool, the corresponding relationbetween the images of the welding pool and the welding parameter was concluded and thecontrast model of the double wire automatic gas shielded arc welding defect through andthe images of the welding pool of aluminum alloy was established. All of the conclusionwould be the elements of the intelligent control based on the visual sensor of the doublewire automatic gas shielded arc welding process.