Detection System for Concentration Quantization of Colloidal-gold Test Strips Based on Embedded and Image Technology

Facing the increasing food safety issues, Chinese government has been carrying out compulsory tests on food to meet the requirements of domestic and foreign markets. Colloidal-gold test strips using the colorimetric principle are widely used for rapid qualitative detection of harmful residues in food. In this research, an embedded system was biult as a solution of quantitative detection of colloidal gold test strips based on image processing technology. Images of the strips were captured and processed by the embedded system which was equipped with an image sensor. The operating system was WINCE 6.0 and the hardware platform was mini2440 based on ARM920T and a CMOS image sensor OV9650. The image processing software developed in Visual C++ included image input, denoise processing, image feature information extraction of effective area, linear transformation for the value of T/C. A functional relation between image feature information of effective area and residue concentration was established comparing to the results of standard chemistry experiments and strip detecting. Sample concentration was obtained by matching the mathematic model with the value of T/C. The results showed that the measurement error of the system was less than 5%.