Research on intelligent transportation and logistics tracking of long billet based on machine vision

We solve the problems in the iron and steel industry of identifying the quantity, length, and position of billets in complex industrial environments by studying and applying machine vision algorithms, such as billet image segmentation based on optimal threshold method and billet image post-processing based on morphology. Compared with the existing technologies that rely on various sensors to realize the automatic transportation of billets or the physical marking of billets one by one to realize the billet logistics tracking, the scheme proposed here has a low one-time investment cost, especially for the billet logistics tracking of follow-up production control and quality traceability. Because the hot delivery billets only carry out virtual marking, there is no consumption cost of consumables, i.e., the higher the hot delivery rate, the lower the average marking cost of billets. Moreover, due to the non-contact of machine vision, the system failure probability is also lower. The machine vision algorithm and solution proposed here achieve the goal of intelligent transportation and logistics tracking of long billets with lower investment cost, lower consumption of consumables, fewer faults, and higher efficiency. (c) 2022 SPIE and IS&T