Research On Embedded HD Electronic Endoscope Image Processing Technology Based On CMOS

Electronic endoscope obtains the body image by CCD or CMOS sensor and transfers the image data by image processor for processing and displaying.As a widely used medical equipment,electronic endoscope can be used for internal lesions diagnosis and surgery.Due to the gap between domestic and foreign technology,most of clinical-used electronic endoscope are still the imported products.The image acquisition and processing system has been designed based on ZedBoard,with CMOS sensor used as photosensitive device.Common image processing functions used in HD electronic endoscope have been implemented in our system.High dynamic range imaging is introduced into the electronic endoscope system and a high dynamic range image fusion algorithm based on image content assessment factors is proposed,which suppresses the ghosting phenomenon effectively.The main work in this paper is as follows:1.The HD electronic endoscope image processing system is designed based on ZedBoard.The acquisition of image signal from CMOS sensor is completed and the real-time display of HD video with frame frequency of 60 fps and resolution of 1920×1080 has been implemented on the system.2.Automatic white balance,automatic exposure and image sharpening algorithms for electronic endoscope are proposed.The influence of noise points is reduced by median filter preprocessing and the influence of over-exposure points the influence of over-exposure points is reduced by the new constraint on the reference white points.An automatic exposure adjustment algorithm based on the combination of mean and peak brightness is proposed and the dynamic step size method is applied to implement the automatic exposure function.A novel image sharpening algorithm is proposed,which is based on the combination of Sobel operator and Laplacian operator.The noise resistance of image sharpening algorithm is improved.3.The HDR image fusion algorithm based on image content assessment factor is proposed and SSIM is used to evaluate the algorithm effect.The experiments results show that the algorithm can suppress the ghosting phenomenon while synthesizing HDR images.