Research On Image Restoration Algorithms In Imaging Detection System

The imaging of objects through atmospheric turbulence is an inevitable problemencountered by optical imaging sensors mounted in missiles or machines working in theturbulence atmospheric environment. Owing to the existence of the atmosphere, beforelight-rays enter the window of the imaging sensors, the atmospheric turbulence willrandomly interfere with the transmission of the light waves came from the objects, causingthe distribution of image intensity values on the focal plane to diffuse, the peak value todecrease, the image to get blurred, and the pixels to deviate, and making objectidentification very difficult. Owing to the fact that the point spread function of turbulenceis unknown, changable with time, and hard to be described by mathematics models, therestoration of turbulence-degraded images is much more difficult and challenging in theworld. This paper concerns the restoration of turbulence-degraded images, focusing theresearch on the new restoration algorithms. Meanwhile, the restoration algorithm ofimages blurred by the high rotational motion of the imaging platform mounted in a missileor a machine is also investigated in this paper. The main contributions of this paper aregiven below. 1. For the problem of restoration of turbulence-degraded images, it is of utmostimportance to make a correct estimation of the turbulence' s stochastic point spreadfunction (PSF). A new method is proposed for estimating the discrete values of theturbulence PSFs, in which two short-exposure images are used as the inputs and a series ofequations for calculating the discrete values of the PSFs are developed. The relationship ofthe stability of the solution for a system of equations and the conditional number of itscoefficient matrix has been analyzed. On the basis of theoretical investigation, someeffective rules for selecting equations have been worked out to ensure a reliable solutionfor the PSFs. In order to overcome the interference of noise, the principle of constraintoptimization has been applied in the algorithm, in which the problem of calculating thePSF values has been boiled down to the least-squares optimum estimation under theconstraint of the PSF values being non-negative and spatial smoothing. Thus, a restorationalgorithm based on the optimum estimation of the PSF values is proposed, which providesa new method for the restoration of turbulence-degraded images. III华 中 科 技 大 学 博 士 学 位 论 文 2. A new restoration algorithm based on anisotropic and nonlinear regularization ispresented for restoring the turbulence-degraded images, in which, in order to estimate thePSF values that are close to the true ones in noisy cases, some a priori knowledge aboutturbulence PSFs are incorporated properly in the process of optimum estimation. Firstly,the constraints of the PSF values being non-negative and spatial smoothing aretransformed mathematically into the penalty terms and added them to the objectivefunction. Secondly, to match the decay nature of turbulence PSFs, an anisotropic andnonlinear regularization function is suggested to regulate the PSF values adequately in theprocess of rebuilding PSF. The PSF values can be estimated by alternating iterativeminimizing the objective function, and the object images can be restored by theconstrained least-squares filtering method. 3. The matrix expression form of wavelet multi-resolution of turbulence-degradedimages is derived from the multi-scale spatial decomposition of signals. A multi-resolutionrestoration method based on wavelet decomposition is presented for the restoration ofturbulence-degraded images. For this method, the turbulence-degraded images are used asthe inputs, for which the multi-scale decomposition will be made by use of wavelettransform. The PSF values in the low-resolution with big scale can be estimated in the lowfrequency subband by using the approximation information. Removing blur andrestora...