Research on Automatic Detection of Microelectrodes Based on Machine Vision Technology

Abstract

Listen

Translate article

Background: Microfabrication has gradually become a hot spot in the manufacturing industry in recent years, and in the process of microfabrication, the processing of microelectrodes is very important. The precision and error control of the microelectrodes play a vital role in the final machining results. Based on digital image processing technology and Matlab software, this paper designs a microelectrode automatic measurement system. Through preprocessing and image segmentation of the microelectrode image, the electrode size can be obtained quickly. Compared with the size measured by the microscope, the detection size of this system has an error of less than 4% and coaxiality measured error within 6%. The results show that the system has the characteristics of high measurement accuracy and fast detection speed and can meet the requirements of rapid size measurement of microelectrodes. Objective: The purpose of this research is to improve the efficiency of size detection when manufacturing a large number of microelectrodes by developing an automatic detection system. This system has made relevant optimization on the error control of the size detection to ensure that the error is within the allowable range. Method: Firstly, the original images of the microelectrodes were obtained using an electron microscope. Secondly, the necessary image processing had been done for the image: gray processing, binarization, median filtering, morphological processing, and large background noise removal. These processes are automatically carried out in the system. In addition, three methods of measuring coaxiality error are proposed, and the first measurement algorithm with the minimum deviation is selected and put into the system, which can ensure the accuracy of coaxiality offset. Finally, the diameter of the microelectrode was measured, and the coaxiality offset of the microelectrode was measured by various methods. After the comparison of the measured values manually, the algorithm with the minimum deviation was selected and put into the system. Results: The diameter and coaxiality of the cylindrical microelectrode and the cylindrical microelectrode with a ball head are successfully measured, and the measurement error is within 4% after comparison, showing that the automatic measurement system has great measurement efficiency. Conclusion: The comparison of the electrode measurement results of this system with the manual measurement results shows that the automatic measurement system developed in this paper can meet the accuracy requirements for measuring the diameter of the microelectrodes, and can significantly improve the measurement efficiency compared with manual measurement.