Research On Detection Of Color Vision Defects And Recoloring For Auxiliary Recognition Methods

Color exists in all aspects of people’s daily life,affecting almost everyone’s life and work.At the same time,color is also a part of many objects.Color can not only simplify the communication between people,but also convey a lot of information.However,color vision deficiency patients account for more than 8% of the global population,and many places have set strict restrictions and corresponding requirements on the occupations that color vision deficiency patients can engage in.Due to the huge number of patients with different types of color vision deficiency and it is difficult to collect,on the one hand,the color data of the color vision test charts currently used in clinics are all collected from patients with color vision deficiency.Single,the accuracy of color vision defect detection is difficult to verify;on the other hand,color vision defect is a genetic disease,which cannot be cured by medical means.Due to the lack of color discrimination ability,color vision defect patients cannot effectively obtain the information transmitted by color,so there are many problems in life.Many inconveniences.In response to the above two aspects,this paper has done two main works:(1)This paper proposes an Ishihara test board generation method based on Monte Carlo method.The theory of color selection was researched,and the color was selected based on three color vision defect confusion lines,and the color vision defect simulation model was used to screen.In the synthesis of the Ishihara test board,the pattern boundary is defined by multiple sets of coordinate points,the large circle with a radius of a certain set value is used as the outer large circle,and the radius of the large circle is scaled and filled based on a certain ratio,and the filled circle points exceed the pattern boundary.The problem is solved by solving the line-circle intersection problem,and setting the corresponding constraints to achieve the ideal filling effect.The position of the filled circle is determined by the crossing number algorithm,and then the selected color is used to fill the circle.Finally,a variety of color vision defect simulation methods are used to compare and analyze the color vision detection map used in traditional clinics.Both subjective and objective evaluations are superior to the traditional color vision detection map,which finally verifies the effectiveness of the method in this paper.(2)A color refill method based on color gamut conversion proposed in this paper.By using the fuzzy clustering algorithm to obtain the confusing color and non-confusing color clustering of the original image,and then projecting the confusion key color to the blue confusion line with the shortest straight line distance to obtain the normal color gamut of the color vision deficiency,and setting the corresponding objective function pair The key color is optimized to ensure the perception of the image by people with color vision deficiency and normal color vision,and then the original image is recolored by using the transmission mechanism between clusters.Finally,the experimental sample results in various scenarios are compared with the other three recoloring algorithms,and a variety of indicators are used for objective evaluation to verify the effectiveness and versatility.