When capturing images, traditional imaging devices can only record images with low dynamic range (LDR) and cannot capture scene information with high dynamic range (HDR) in the real world due to their own response characteristics. Therefore, high dynamic range imaging technology has developed rapidly and gradually become a research hotspot in the field of digital image processing. The main problems of current multi-exposure images to generate HDR images include ghosting caused by camera shake and foreground objects moving, and information loss in overexposure and underexposure situations. To solve the ghost problem, the gradient direction histogram descriptor is used to process multi exposure images, and the brightness invariant motion estimation technology based on feature optical flow is used to generate high dynamic range images. The algorithm first selects the well exposed image as the reference image, and uses the gradient direction histogram descriptor which is robust to the illumination change to process the images with different exposure times, then uses the optical flow estimation algorithm to distort the input image to the reference image according to the corresponding relationship between image features ,finally synthesizes the final HDR image according to the weight map .The experimental results show that compared with the existing HDR algorithms, the proposed method has a certain improvement in performance.
Compressed sensing theory is a new sampling theory, which provides a method to recover the original signal from a small number of samples. For sparse signal and compressible signal, compressed sensing theory compresses the signal while sampling. It combines the sampling process and compression process. It breaks through the traditional Nyquist sampling law and saves a lot of storage, transmission, computing and other resources. This theory not only reduces the cost of storage and transmission of digital image and video acquisition, but also provides a new opportunity for the follow-up research of image processing and recognition, and promotes the combination of theory and engineering application. It includes three parts: sparse representation of target, design of measurement matrix and reconstruction of target. Reconstruction algorithm is a key step in the process of compression imaging, which determines the accuracy and speed of image reconstruction to a certain extent, so it is very important to select the appropriate image quality evaluation index. The image quality evaluation of existing reconstruction algorithms mainly focuses on peak signal-to-noise ratio (PSNR) and structural similarity (SSIM). The advantages of these two algorithms lie in simple algorithm, fast inspection speed, which are suitable for evaluating the advantages and disadvantages of algorithms, but the disadvantage is that they can only be evaluated on the basis of known original images. In the actual imaging process of compressed sensing, it is impossible to obtain the original image, so we need to use an image quality evaluation method which is not based on the original image.
Images store a lot of information and are the window for human beings to understand things. A lot of research is devoted to analyzing and processing images, which is called image processing in a broad sense. Image processing includes image recognition, image restoration, image enhancement, image coding and so on. This paper mainly focuses on the field of image restoration. Image restoration, also known as image inverse problem, aims to restore high-quality original images from degraded or damaged observations. It also acts as a preprocessing step in many intermediate and advanced image processing tasks. Due to the limitations of sensors or environmental conditions, imaging systems usually have factors such as noise, optical or motion blur, resulting in image degradation and distortion. Aiming at the ill posed problem of image pixel missing and blur in the process of compression coding, this paper uses GMM model to solve the degraded image, so as to achieve the purpose of image restoration.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.