Research On Detection System Of Image-based Fluorescence Labeling

At present, fluorescent strip based on the fluorescence labeling technology hasbeen widely used in detecting target object, such as the antigen and antibody. On thefluorescent strip, the colour depth of tested area reflects the concentration level oftested substance. The current detection method is usually through eyes to give a roughestimate, so it is only qualitative measurement, instead of quantitative measurement.This detection method does not only have relative low detection sensitivity, but alsohas a high requirement of the fluorescence intensity.This thesis aims to study the combination of virtual instrument technology andfluorescence detection technology that is applied to the fluorescent labeling stripimage detection. And then, this thesis designs a fluorescence labeling detectionsystem based on virtual instrument. On the one hand, the system takes theLabVIEW8.6as the operating platform, which designs the hardware part that baseson the excitation light and image acquisition. This system adopts the IMAQ USBdevice instead of the traditional image acquisition card, which not only can reduce thecosts of the system development, but also make the operation method more simpleand efficient. Finally the system realizes a higher and more reliable automationdegree. On the other hand, this system design the software part that focuses on theimage acquisition, denoising and analysis to achieve the detection of fluorescencelabeling image.This thesis takes the ractopamine solution as an example to carry on the detailedexperiment.Firstly,this thesis makes use of the image fluorescent detection systemdesigned to collect strips fluorescence images, and then to carry on the filtering anddenoising. Eventually, the system achieves a quantitative analysis of the targetfluorescence image with the numerical values and the image forms. The system givesthe peak value of the image, a half-width of the peak values, the area values and theconcentration ratio. The experiments prove that this system has good accuracy and stability. In addition, the experiment result data can be stored through the outputdevice, which is convenient for query and statistical analysis in the future.