Research on the detection of gold immunochromatographic assay by the image histogram feature vectors and fuzzy C-means

The gold immunochromatographic assay has the advantages of simple operation, low costs and rapid operation time. But the traditional immunochromatographic strip can only get qualitative or semi-quantitative results observed directly with the naked eyes, the disadvantages are low measurement accuracy, and it is difficult to achieve quantitative measurement. This paper presents a new method to perform the unsupervised classification for the gold immunochromatographic strip by combining the histogram features vectors and the fuzzy C-means algorithm based on computer image analysis system. Then provide procedures of extracting the histogram features as input vectors for fuzzy C-means algorithm and example of the fuzzy C-means clustering analysis methods of the gold immunochromatographic strip. In the experiment, the discrimination coefficient Classification coefficient F is 0.8953 and the Average fuzzy entropy H=0.085, the gold immunochromatographic strips with various human chorionic gonadotropin (hCG) concentrations were accurate and unsupervised classified to three clustering. The result proves that the classification of the gold immunochromatographic strips by the histogram features vectors and the fuzzy C-means algorithm is reasonable and validated, it offers a good semiquantitative and quantitative test method to the immunochromatographic strip for clinical diagnosis. The research can not only enhance the detection sensitivity and the objectivity of test result, but also have a sound application value.