The Research Of Color Space Fusion Color Difference Method In Milk Bacterial Content Detection

Recent years,security incidents of milk products occur frequently,milk bacterial content detection has been more attention as one important index of milk products.The traditional detecting milk bacterial content is mainly the plate count method,even though the detecting accuracy is high,it takes 48 hours or more.Some other methods based on the plate count have been improved,but still need 12 to 24 hours.Some of the latest detecting methods in domestic and abroad such as flow cytometry,ATP bioluminescence method,significantly reduced detecting time,but instruments and supplies are expensive and difficult to achieve large-scale application.In order to balance the need for rapid detection and application,this paper firstly analyzed the proliferation of bacteria in milk,based on the detection of milk bacteria content in color difference method,proposed the multiple color space feature fusion theory applying to color difference method,improved the color difference resolution.In the part of detection,original mode has been changed,which only received the reflected light by a single channel.A "liquid color detecting device" was designed by using dual transmission method.The device comprises a D65 light source,cuvettes,color sensors,a main controller,a LCD screen etc.And through the white balance and other software design achieved the overall system optimization and controlled.Experimental results have shown that this measuring device has a good stability,and can detect the color difference values of solution in real time,finally gives the bacterial content of milk.Combined with the use of multiple color space feature fusion color difference method,it can be more accurately to determine the small color difference when similar color appeared.Detecting time is shortened to severaltens minutes to 6 hours.It is easy to operate and maintain with low cost.To achieve more accurate detection,a large number of samples need to be collected,more accurate mathematical relationships can be established with more data,and the color detecting device needs further improvement.