Automatic recognition method for optical measuring instruments based on machine vision

Abstract

Based on a comprehensive study of various algorithms, the automatic recognition of traditional ocular optical measuring instruments is realized. Taking a universal tools microscope(UTM) lens view image as an example, a 2-layer automatic recognition model for data reading is established after adopting a series of pre-processing algorithms. This model is an optimal combination of the correlation-based template matching method and a concurrent back propagation(BP) neural network. Multiple complementary feature extraction is used in generating the eigenvectors of the concurrent network. In order to improve fault-tolerance capacity, rotation invariant features based on Zernike moments are extracted from digit characters and a 4-dimensional group of the outline features is also obtained. Moreover, the operating time and reading accuracy can be adjusted dynamically by setting the threshold value. The experimental result indicates that the newly developed algorithm has optimal recognition precision and working speed. The average reading ratio can achieve 97.23%. The recognition method can automatically obtain the results of optical measuring instruments rapidly and stably without modifying their original structure, which meets the application requirements.

Keywords

automatic recognition / optical measuring instruments / template matching / neural network

References

[1] Chen Z., Lin Y., Zhao M., et al. Pretreatment algorithm of numeric character auto recognition in view field of eye lens on universal tools microscope[J]. Journal of Optoelectronics Lasers, 2005, 16(1): 80-82.

[2] Shi M., Fujisawa Y.. Handwritten numeral recognition using gradient and curvature of gray scale image[J]. Pattern Recognition, 2002, 35(10): 2051-2059.

[3] Wang S., Zhao M., Lin Y.. Design of a portable auto-reading system based on DSP of universal tools microscope[J]. Modular Machine Tool and Automatic Manufacturing Technique, 2006, 3(3): 70-72.

[4] Stefan K., Leon P.. Handwritten digit recognition by neural networks with single-layer training[J]. IEEE Trans on Neural Networks, 1992, 3(6): 962-968.

[5] Perantonis S., Lisboa P.. Translation, rotation, and scale invariant pattern recognition by high-order neural networks and moment classifiers[J]. IEEE Trans on Neural Networks, 1992, 3(2): 241-251.

[6] Raicharoen T, Lursinsap C. Critical support vector machine without kernel function [C]. In: Proceedings of the 9th International Conference on Neural Information Processing, IEEE 2002. 2532–2536.

[7] Delopoulos A., Tirakis A., Kollias S.. Invariant image classification using triple-correlation based neural networks[J]. IEEE Trans on Neural Networks, 1994, 5(3): 392-409.

[8] Chen X., Pei X.. The Technology and Application of Artificial Neural Network[M]. 2003, Beijing: China Electric Power Press.

[9] Khotanzad A., Hong Y. H.. Invariant image recognition by Zernike moments[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 1990, 12(5): 489-498.

[10] Teh C. H., Chin R. T.. On image analysis by the methods of moments[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 1988, 10(4): 496-513.

[11] Wang H., Chen Y., Huang Y.. The technology of virtual instrument based on PC[J]. Metrology and Measurement Technique, 2005, 32(6): 32-34.