Machine Vision Based Precise Detection Of PCB Micro-Drill's Geometry Parameters

In order to adapt to the trend of PCB micro-drills becoming smaller and smaller (mainstream products in the diameter of 0.2mm, the smallest is 0.02mm), research and development on the high-speed, high-precision, highly integration PCB micro-drill's detection technology and equipment has positive social and economic benefits and broad application prospects. Machine vision based PCB micro-drill precision detection equipment adapts image processing technology,automatic control technology,virtual instrument technology,mechanical technology and integrates factors of mechanical,electrical,optical,gas into one machine, which has superiority than other detection equipments. This dissertation studies how to achieve precision detection of PCB micro-drill from mechanical structure,vision system and image processing, based on one self-developed PCB micro-drill precision detection equipment.Mechanical structure is foundation of the whole detection system. This dissertation discusses how to ensure precision detection from the hardware foundations such as equipment overall structure, feeding and landing structure, feeding disk structure, good product and bad product disk structure, rotation and height adjustment structure. Vision system is critical to the whole detection system. This dissertation designs vision systems of PCB micro-drill's flank and first facet according to their differences, including the selection of cameras,lens,lights. Image processing technology influences the detection precision and speed. This dissertation studies the detection algorithms of PCB micro-drill's flank and first facet according to their detection items, focusing on the adaptive contour corner detection method of micro-drill's first facet. This dissertation also analyzes and adjusts several factors that affect the accuracy of detection. Study of the equipment's HMI is provided finally.Experimental results of PCB micro-drill's geometry precise detection show that detection precision is under 1μm and detection speed can reach 20/min. The detection technology studied in this dissertation provides a powerful technical assurance for the practical application of PCB micro-drill's geometry detection equipment.