Study On Detectiong Method For Smoked Meat Based On Computer Vision Technology

Food safety issues is related to people’s livelihood strategy. Smoked meat with its uniqueflavor is favored by consumers, and Qiqihar’s smoked meat is famous nationwide. But themeat can produce toxic or harmful substances acrylamide while it is being smoked on hotcoals. The method of determining acrylamide contents in food has been listed as China’s "15"national key "key to food safety technology" and "The Eleventh Five-Year" national scientificand technological supporting major projects planned "food processing key technologyresearch and industrial development" research topics, and there are many achievements, suchas the meteorological chromatography, liquid chromatography, gas chromatography massspectrometry and other chemical detection methods. But these methods have somedisadvantages, for example, destroying the sample, expensive detecting equipment, longdetecting time, complicated detecting process.In response to these problems, we propose a determining method of smoked meatacrylamide content value based on computer vision. Using the proposed algorithm of imagefiltering, algorithms of image segmentation and color measurement to measure smoked meatdouble surface color values. Analysis of the resulting surface color values and their smokedmeat acrylamide content value, a strong correlation between smoked meat double surfacecolor value and acrylamide content is found, meanwhile, the fitted linear regression equationis given. When acrylimade content is determined, put the doubl surface color value a intowell-established linear regression equation, and then the acrylamide content can be calculated.Experimental results show that the proposed method achieved the purpose that it couldquickly and accurately determine in smoked meat without destroying the sample.