Rapid multi-resolution denoising method and device under hybrid noise model

Abstract

The invention discloses a rapid multi-resolution denoising method and device under a hybrid noise model. The method comprises the following steps: after acquiring an image with noise, building a multilayer gaussian pyramid and a multilayer Laplacian pyramid corresponding to theimage with the noise; performing a plurality of times of iteration calculation to obtain total variation regularization denoising target functions respectively by using a gradient descent method; performing the final iteration calculation to obtain a denoised image and the like. The device comprises a storage for storingat least one program, and a processor for loading at least one program for implementing the method. By adopting the rapid multi-resolution denoising method and device, the plurality of times of iteration calculations of the gaussian pyramid and the Laplacian pyramid are combined, the target functions can be solved corresponding to different image resolutions in each iteration calculation, the initial value of each iteration calculation is derived from the result of the previous iteration calculation and the Laplacian pyramid, denoising effect can be ensured, the calculation load of resolvingis lowered, the resolving speed is increased, and the calculation efficiency is increased.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a multi-resolution fast denoising method and device under a mixed noise model. Background technique [0002] In the industrial field, as a non-destructive, non-contact and high-resolution defect detection method, X-ray image detection is an important means of integrated circuit defect detection. It can detect solder joints, wires and Defects such as pressure welding between wires, cracks, holes and bubbles in the circuit can effectively solve the problem of difficult detection of internal defects of packaged components in the process of packaging integrated circuits, and ensure the production quality of integrated circuit packaging. Great application value and economic benefits. In the medical field, X-rays are also often used to detect lesions. X-ray detection is to use X-rays to penetrate the object and receive it by the image receiving device, and record the intensi...

AI Technical Summary

Problems solved by technology

[0012] From the above process, it can be seen that the data fidelity term of the full variation regularization objective function under the Gaussian-Poisson mixed noise model is seriously nonlinear, and the existing denoising methods based on the full variation regularization objective function must iterate repeatedly , and each iteration must include two processes of outer iteration and inner iteration, and each iteration must update s t with η t Two coefficients, at u t+1 Fixed s in the optimization objective function of t with η t After that, we can solve u t+1 The inner iteration of the algorithm also needs to perform complex two-dimensional matrix calculations, so the number of iterations required is large, the calculation amount is large, the calculation speed is slow, and the algorithm operation efficiency is low

Even if the original noisy image is directly down-sampled to reduce the image resolution, and then iteratively solved by the corresponding full variation regularization denoising method, the overall running time of the algorithm can be reduced and the solution efficiency can be improved. When this method is used, the key edge information of the image has been lost during downsampling, so that after the subsequent full variation regularization denoising method is used and solved, the edge area in the denoised image is obviously blurred, and it cannot be obtained. Satisfactory denoising effect

Images