The Study Of Quantitative Detection Of Fluorescence Reagent Image Based On FPGA

Immune chromatography technology can detect specific biochemical samples quantitatively via the analysis of excitation spectra and fluorescence spectra from fluorescent markers.With many attracting characteristics such as high detection accuracy,fast detection speed,simple and convenient operation,etc.,this technology is widely used in medical detection,food safety and environmental protection.However,domestic and foreign production of fluorescent immune chromatography quantitative detection instrument presently is mainly based on photoelectric detection method.Certainly,this method need high precision positioning of the test strips,and a transmission device is required.If the detect position is not accurately addressed,the detection results will be affected.In addition,the transmission device will cause the mechanical vibration noise which will affect the quantitative detection result.Whereas,image processing method does not need accurate positioning and transmission device.As the FPGA(field programmable gate array)inherits the advantages of ASIC(integrated circuit)and with processing speed over the DSP and microprocessor,FPGA is widely used in a variety of real-time image processing system in recent years.Therefore,this paper chooses FPGA as fluorescent images of fixed quantity detection processing core.In this paper,the main content of this paper is described as follows:(1)This part demonstrates the principle of quantitative detection for the fluorescence immunoassay and illustrates structure of the test strip,followed by introducing the mathematical theory of quantitative detection for fluorescence immunoassay.Finally,the design of the device used in this paper is proposed.(2)Based on the quantitative detection principle in part(1),the fluorescence images collected by image acquisition device are analyzed and biochemical noise is explained.On the basis of this select pre-segmentation algorithm and enhancement algorithm which are suitable for FPGA implementation,image enhancement processing and the plane fitting method is performed to remove the noise of gray drift.Finally,segmentation algorithm is performed to extract test line and control line.Besides,quantitative detection algorithm is implemented.(3)Based on the characteristics of image processing algorithm,quantitative detection algorithm and structural features of the FPGA,algorithms are optimized and implemented on FPGA platform,which improves the execution speed of these algorithms in FPGA.Finally,testbench and the corresponding timing simulation is conducted to verify the feasibility of the algorithm.(4)This part introduces the process of extracting characteristic value of quantitative detection,and analysis influence of the different feature extraction method on detection repeatability and accuracy influence.Finally,the eigenvalue curve fitting is determined,and ESE Quant lateral flow reader(QIAGEN,Germany)is chosen to analyze the detection results as comparison.Experimental results show that quantitative detection using this proposed processing method,CV of the characteristic value(sum of the pixels)is less than 1%.Besides,characteristic value of the fitting curve for quantitative detection has a linear correlation coefficient greater than 99%.Furthermore,measurement range of the detection system is wider than the ESE by adjusting the exposure time of the camera.