Color Correction And Recognition Method Of Test Paper Based On Perfect Reflection And Gray World

Changes in the composition and content of urine can largely reflect the health of human urinary system and kidney.Urine dry chemistry test paper is a commonly used method to get urine component content by changing the color of test paper after urine reaction.Finding a way to monitor human urine composition and content in a timely,economic and convenient way has become an urgent problem.With the development and popularization of the mobile phone,the color of the test paper is automatically identified by the computer and the color of the test paper is automatically recognized by the computer,thus the content of the urine composition is obtained,and the purpose of monitoring the health of the human body is achieved.But the test paper image in the use of mobile phone will be affected by the impact of ambient light,affect the color recognition results,make the final content of urine content is inaccurate,the human health of the judgment is wrong.In this paper,the problems existing in color recognition of test paper in urine test paper are studied in the aspects of color correction and recognition algorithm of test paper.The following work is carried out in the aspects of image color detection,color correction,main color extraction and color recognition.(1)According to the color distribution characteristics of the color test paper in Lab space,the color detection algorithm of the test paper is studied.First,the basic information of the white correction block is obtained under the standard light,and the ratio of the first moment and the two moment of the white correction block under the standard light of the white correction block under the standard light under the unknown light is obtained.The partial color factor K of the color of the test paper is determined.According to the size of the K value,the color degree of the test paper is judged.The larger the K value is,the more serious the color of the test paper is.Experiments show that the algorithm can effectively identify the color test paper.(2)Combining the theory of perfect reflection and gray world,the color correction algorithm of test paper is studied.First,the split test paper image and the white correction block image are obtained.Then the maximum value and the average value of each RGB channel are calculated according to the perfect reflection,and the white correction block image is calculated according to the gray world.The statistical mean value and the total mean value of the RGB are made up of the two element first equation group,and the channel adjustment coefficient of the RGB channels is obtained.Finally,the square adjustment of each pixel's RGB value is made according to the adjustment formula of each channel.Experiments show that this algorithm has the best color correction effect for test paper images,and can effectively reduce the color bias of test paper images.(3)In view of the color distribution of only one main color on the test paper strip,the main color extraction algorithm of the reagent block image is studied.First,the main color 1 is calculated with the average value method,then the color histogram of the reagent block image is obtained and the main color 2 is calculated by the maximum frequency method.The color value of all pixels and two kinds of pixels are measured by the standard deviation.The main color deviation is calculated and the nearest main color is chosen.Experiments show that the main color extracted by this algorithm is the closest to all pixel values.(4)Combining with the color process of human eye recognition test paper,the color recognition method of test paper is studied.The standard color of the main color and the concentration level of I extracted from the reagent block picture is regarded as two color vectors,and the angle between the two color vectors is calculated and the minimum angle of the angle is selected as the final color recognition result.Taking human eye recognition results as a reference,we evaluate the effectiveness of the recognition results.The experiment shows that the accuracy of the test paper color recognition is high.The innovation in this paper can be described as below:(1)The algorithm of image color detection based on Lab color space is improved.(2)The color correction algorithm based on perfect reflection and gray world is improved.(3)The primary color extraction algorithm based on standard deviation is proposed.(4)A color identification method based on color difference of space angle is proposed.