Improvement Design Of Image Processing And Packing Problem For Punching Auto-feeding Machine

Punching machines are widely used for industrial production thanks to its performance and energy-saving properties.The traditional production method is manual and thus relatively dangerous and inefficient.Punching auto-feeding machines change this situation to a large extent,but it also accounts for many of the defects,such as,the existence of errors in image processing,the inefficiency of work piece packing problem and the overly high cost,etc.The paper aims at resolving these problems above.First,the previous punching auto-feeding machine acquires the camera working plane using mark points during runtime.In this paper,a new solution is conducted,where the camera working plane is determined during the installation of the machine.Second,the research distinguishes different material types,like,bar-type,arbitrary type,etc.The idea is to divide the whole problem and conquer each part using a proper algorithm.Third,the paper designs an embedded machine vision system,which replaces the usage of industrial camera.The solutions in the paper are strictly tested both in test and production environment,which achieves the objective of the paper.