An automatic chip structure optical inspection system for electronic components

An automatic chip structure inspection system based on machine vision is presented to ensure the reliability of electronic components. It consists of four major modules, including a metallographic microscope, a Gigabit Ethernet high-resolution camera, a control system and a high performance computer. An auto-focusing technique is presented to solve the problem that the chip surface is not on the same focusing surface under the high magnification of the microscope. A panoramic high-resolution image stitching algorithm is adopted to deal with the contradiction between resolution and field of view, caused by different sizes of electronic components. In addition, we establish a database to storage and callback appropriate parameters to ensure the consistency of chip images of electronic components with the same model. We use image change detection technology to realize the detection of chip images of electronic components. The system can achieve high-resolution imaging for chips of electronic components with various sizes, and clearly imaging for the surface of chip with different horizontal and standardized imaging for ones with the same model, and can recognize chip defects.