Color-blind Image Processing System Design And Algorithm Study

Color blindness and color weakness is a common sex-linked genetic disease which is mainly caused by the variation or missing of cone cells in the retina. According to statistics, about 8% of the male are suffering from color blindness or color weakness, and the ratio is about 0.5% for the female. Color blindness brings much inconvenience in the life, studying, working and so on. To decrease the inconvenience and improve the ability of color discrimination for dichromats, some studies have been done in the fields of medicine, biology, psychology and informatics. In the field of informatics, re-coloring images for dichromats is noninvasive and more feasible to help them.In the study of color-blind image processing, the hardware system design is a very important part which needs high speed and accuracy. The embedded system has some good characteristics, such as flexibility, software and hardware modifiable, mini-bulk, low cost, and low power. Among the 32-bit embedded systems, ARM (Advanced RISC Machine) is the accredited leading processor. In this paper, an ARM-based embedded system design scheme is proposed for the color-blind image processing. The hardware and software of the embedded color-blind image processing system are designed using an ARM core processor. Besides, a simple and convenient interface is implemented. This system, which supplies a general hardware platform for the applications of color-blind image processing algorithms, may bring convenience for the investigation on both color blindness and its rectification.Besides, according to the speed and accuracy requirements of practical application, a color blindness rectification method is proposed based on the geometric transformation mapping of images. Based on the color ratio on both sides of the specific plane in the color space, each plane is divided into different mapping regions by geometric transformation. Then the colors of the image are transformed into the ones which dichromats can distinguish. This method can improve the color discrimination ability of dichromats. Meanwhile, it is fast enough for real-time applications.