• DocumentCode
    3202793
  • Title

    Research on Color Recognition of Urine Test Paper Based on Learning Vector Quantization (LVQ)

  • Author

    Wang Chunhong ; Zhang Hongqiang ; Yu Changxing

  • Author_Institution
    Coll. of Electr. Eng., Suihua Univ. Suihua, Suihua, China
  • fYear
    2012
  • fDate
    8-10 Dec. 2012
  • Firstpage
    850
  • Lastpage
    853
  • Abstract
    A color recognition of urine test paper based on learning vector quantization (LVQ) is proposed. According to the complicated nonlinear relationship of color space transformations, acquire the standard color threshold value. After the normalization processing, establish the color recognition model of urine samples based on LVQ neural network. The results indicate that color recognition of urine samples is feasible and effective by the way of LVQ neural network, compared with the method of color chromatic aberration evaluation, this method could omit the step of color space conversion and is easy to understand and operate, meanwhile, only with the original equipment RGB and color space RGB values, the prediction can be brought about, it is completely necessary to normalize the data with LVQ network in classification processing.
  • Keywords
    image colour analysis; image recognition; learning (artificial intelligence); medical image processing; neural nets; LVQ neural network; color chromatic aberration evaluation method; color image recognition model; color space RGB values; color space transformations; complicated nonlinear relationship; learning vector quantization; normalization processing; standard color threshold value; urine test paper; Biochemical analysis; Biological neural networks; Image color analysis; Neurons; Standards; Training; Learning vector quantization (LVQ); biochemical analysis; color recognition;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Instrumentation, Measurement, Computer, Communication and Control (IMCCC), 2012 Second International Conference on
  • Conference_Location
    Harbin
  • Print_ISBN
    978-1-4673-5034-1
  • Type

    conf

  • DOI
    10.1109/IMCCC.2012.205
  • Filename
    6429040