Real-Time Visual Inspection and Rejection Machine for Bullet Production

摘要

The aim of this work is to use automatic visual inspection and rejection system to improve quality control of bullet cartridges M193 during manufacturing processes. This system will replace the current procedures of human inspection. The automatic visual inspection system consisted of a bullet feeder and high-resolution cameras to capture images of moving bullets for digital image processing. The captured images will be analyzed in real-time to determine any possible defect from dark spots, crack lines, deformation, uneven color, dimension errors, and etc. The image analysis software based on Lab VIEW inspected 2 bullets per second and sent command to reject defect bullets using a PLC controller. This automatic visual inspection system allowed inspection with much faster speed and higher accuracy. The results showed that our system can detect defective bullet with 86.7% accuracy. In addition, the database of different defect types will be useful for determining and eliminating causes of defect.