Research Of Image Enhancement Algorithm Of Retinex And Implementation By Using FPGA

Railway is the artery of national economy, the national infrastructure andthe public transportation device. A safe railway transportation system is criticalnot only for its survival and development but also for the economic developmentand social stability of the country.Trouble of moving Freight car Detection System uses images as majorinformation source. Information included in the images is the importantevidence for manual and automatic recognition of the train’s operation status. Itis also the major evidence for predicting the train’s potential problems. Thus,the quality of the images directly influences the effectiveness and safety of therailway operation system.Retinex is a novel digital image processing algorithm, which separates theincident component and reflection component of an original image, therebyobtaining an objective’s reflection component without the influence of theincident component. This system mimics human brain and eye to acquire andprocess an image, thereby enhancing the image quality. It is very effective inremoving the halos in an image.In this paper, we assessed a series of traditional image-enhancing algorithm,compared their image-processing effectiveness under specific conditions, andextensively studied the effect of Retinex on TFDS system. Based on Retinex, anonlinear expanding of the image was processed, which significantly enhancedthe image quality. The major disadvantage of this system is time-consuming andslow when it runs in regular computer. Thus, it cannot satisfy the actual needs.After a comprehensive comparison of hardware implementation, we used “imageenhancement based on FPGA” technology to perform and satisfy the TFDSsystem. A detailed description for the transplantation and simulation in theprocess of algorithm transplantation as well as how the hardware system wasdesigned is included in this paper.