Study Of Immunochromatographic Quantitative Detection Based On Image Acquisition Device

Immunochromatographic detection technology is the trend of post-of-care test (POCT) at present. The advantages of the technology such as accurate detection result, ease of use, rapid detection, needless of professional guidance make the application widely. But immunochromatographic test is mostly qualitative or semi-qualitative confined to the human eyes, which limit its further application and promotion. Immunochromatographic quantitative detection is the inevitable trend of development of the technology. Immunochromatographic quantitative detection methods include photoelectric reflex method and image processing method. Existing photoelectric reflex method of immunochromatographic quantitative detection has good reproducibility among units and accuracy. However, high precision is required for the strip position in this method. As less of verification among units, the deviation among units is difficult to be reduced.Strip transmission unit is needless in image processing method, strip positioning can be eliminated from software. Information of color image is richer and superior to improve the precision. In this paper, a color image acquisition device is designed, immunochromatographic strip image is processed in the color space to achieve quantitative detection. Image detection method is used to make the verification strip of photoelectric reflex method in quantitative detection, and provide a method to verify the quantitative detection results by using photoelectric reflex method. The main tasks of this paper includes the following sections:(1) Mathematical model quantitative detection, A high stability of parallel white light source combined with CMOS image sensor with high SNR and low power consume is used to designed immunochromatographic strip color image acquisition device with high color reproduction.(2) Analysising the hue of immunochromatographic strip color images. The primary hue of the strip image areas (strip shell, background area, test line) are different, but the pixels of regional boundary belong to which region presence ambiguity. The HSV color space is consistent with human visual system and it is selected as the color space for color image segmentation. Using the hue histogram peak point of the strip as FCM algorithm clustering center to segment the color images, the test line is accurately extracted, and the algorithm improved the computational efficiency.Analysis the characteristics of the color spaces. The color components of HSV color space are weakly related and the HSV color space is selected as the feature space. Eleven different concentration HCG solution samples are configuration to carry out experiment. Sample the strip image at different time and extract the test lines. Compared the linear correlations of HCG concentration and eigenvalue at different image acquisition time and different dimensions. The average values of H in test line is used as eigenvalue and the maximum linear correlation of HCG concentration and eigenvalue reach 95.24% at 14 minutes.Valley values of spectral reflectance in the test line area are used to validation quantitative detection accuracy of color image processing method. The correlation coefficient between spectral reflectance Valley values and eigenvalues is reach 98.97%. Determine the test strip image information of the different OD values by color image processing method. And manufacture verification strip for quantitative detection of photoelectric reflex method. The correlation coefficient between the spectral reflectance valley values and OD values of verification strip is 99.47%.In this paper, image acquisition system is designed, the test line is extracted by improved FCM algorithm, realizes a immunochromatography quantitative detection in HSV color space. And manufacture verification strips for photoelectric reflex quantitative detection in this way. It is of great theoretical significance and practical value.