Research Of The HDR Design Based On Multi-frame Exposure Images

The dynamic range of an image refers to the ratio of the brightness of the brightest part to the darkest part of the image.For an image obtained from a scene with a larger brightness range,since the dynamic range of an ordinary color image is much smaller than that of a real scene in nature,highlight and dark areas will appear after the human eye perceives them.High dynamic range technology can solve this problem well.After HDR processing,the highlights and dark areas can obtain better details and levels than ordinary images,which is closer to the real visual effect.With the development of HDR technology,it has been widely used in the fields of security monitoring,medical imaging,consumer display equipment,and telemetry and remote sensing.At present,HDR technology software algorithms are mainly divided into two types: The first is to achieve the purpose of expanding the dynamic range by combining multiple frames of images with different exposures according to the fusion algorithm;the second is based on a single frame of exposure image,through a variety of algorithms Realize HDR images.The HDR image generation algorithm based on a single frame exposure image has lower requirements for the image,but because the image has less single details,the generated image quality is average;the HDR image generation based on the multi-exposure image involves the fusion of multiple frames,although there is The problem of ghosting and image registration,but because the fusion of the image has rich details and better quality,more and more researchers pay more and more attention.The traditional algorithm takes a long time due to the complicated calculation process,which cannot meet the requirements of real-time processing of HDR video,and the development process is based on a software platform,which often has errors with the hardware system.Some scholars at home and abroad have realized HDR video processing based on FPGA platform by improving traditional algorithms,but the video resolution or frame rate is low.Based on the application requirements of ISP,thesis proposes an HDR video algorithm based on multi-frame exposure images and easy to implement by hardware.This algorithm performs pixel correction and ghost removal on two frames of source images according to the exposure and movement of the pixels,and then according to the calculated weight coefficients Complete image fusion,and evaluate the image results in various aspects.According to the algorithm,thesis proposes a hardware implementation scheme,designs the overall architecture and divides the functions of each module,carries on the assembly line design to the module,elaborates the control signal in detail,carries on the RTL simulation verification and the comprehensive realization to the design.The simulation results show that in comparison with the two classic algorithms,the HDR image processed by the algorithm in thesis has a better effect,subjective and objective evaluations are higher,and there is no ghosting phenomenon in the moving area.Both the overexposure area and the underexposure area have been enhanced in detail,the dynamic range of the image has been improved,and the hardware simulation results are consistent with the algorithm simulation results.Comprehensive results and FPGA test results show that the hardware design in thesis can meet the real-time processing requirements of HDR function of 4K,60 fps high resolution and high frame rate video.