论文信息 - The comparison of method for numeral recognition algorithm

The comparison of method for numeral recognition algorithm

On the basis of further research in the numeral image recognition algorithm, this paper puts forward a numeral image recognition based on MUC which is compared with a image recognition algorithm based on computer, both algorithms are by extracted feature information of figure. Experiments show that, the average recognition rate of numeral image recognition algorithm based on MCU can arrive at 74%, and the other based on computer arrives at 96.5%.But numeral image recognition algorithm based on MUC is simple, easy to realize, has low cost and a great potential for growth.

Bin Ma | Zhonghua Han | Kuan Huang | Changtao Wang | Shuhai Bian

[1] H. Genchi,et al. Recognition of handwritten numerical characters for automatic letter sorting , 1968 .

[2] Meng Qingbin. Design of Digital Image Acquisition System Based on MCU , 2011 .

[3] Majid Ahmadi,et al. A hierarchical neural network architecture for handwritten numeral recognition , 1997, Pattern Recognit..

[4] Wang Guang-wei. A Research of Thresholding Technology on Digital Instrument Image , 2011 .

[5] Michael W. Berns,et al. Digital Image Processing and Analysis , 1986 .

[6] Kang Lingzhou. Image Capture and Preprocessing System Based on FPGA , 2011 .

[7] Li Lian-shen. On Handwritten Numeral Recognition Based on Physical Characteristics , 2010 .

[8] Meng Qingbin. New Algorithm of Numeral Image Recognition Based on Feature Information , 2011 .

[9] Fumitaka Kimura,et al. Handwritten numerical recognition based on multiple algorithms , 1991, Pattern Recognit..