Edge-preserving Interpolation Method Research Articles

Ray Tracing Acceleration Using Total Variation Norm Minimization for Radio Map Simulation

In this letter, a ray tracing (RT) acceleration method based on the total variation norm minimization (TVNM) scheme is presented. RT is a general tool used to analyze wireless communication environments. However, the simulation is time consuming because of the large number of ray calculations. This letter focuses on radio map interpolation as an acceleration approach. For radio map interpolation, we adopt TVNM, which is a representative edge-preserving interpolation method in the field of image processing. We regard the non-line-of-sight and line-of-sight region boundary as the edge of the radio map and use TVNM to achieve accurate interpolation. Furthermore, we consider known position selection to enhance the interpolation accuracy of TVNM. A computational simulation using commercial RT software revealed that the interpolation accuracy of the proposed method was better than those of other radio map interpolation methods and that RT simulation can be accelerated with an acceptable interpolation error.

Coarse-to-Fine PatchMatch for Dense Correspondence

Although the matching technique has been studied in various areas of computer vision for decades, efficient dense correspondence remains an open problem. In this paper, we present a simple but powerful matching method that works in a coarse-to-fine scheme for optical flow and stereo matching. Inspired by the nearest neighbor field (NNF) algorithms, our approach, called coarse-to-fine PatchMatch, blends an efficient random search strategy with the coarse-to-fine scheme for efficient dense correspondence. Unlike existing NNF techniques, which are efficient but yield results that are often too noisy because of a lack of global regularization, we propose a propagation step involving a constrained random search radius between adjacent levels of a hierarchical architecture. The resulting correspondence has a built-in smoothing effect, making it more suited to dense correspondence than the NNF techniques. Furthermore, our approach can also capture tiny structures with large motions, which is a problem for traditional coarse-to-fine methods. Interpolated using an edge-preserving interpolation method, our method outperforms the state-of-the-art optical flow methods on the MPI-Sintel and KITTI data sets and is much faster than competing methods.

SU‐E‐I‐62: A New Metal Artifact Reduction Algorithm Using Edge Preserving Smoothing Filter

Purpose: Metallic materials such as tooth supplement or surgical clip lead to metal artifacts on the X‐ray computed tomographic (CT) image. In this study, we have developed a new metal artifact reduction algorithm. To improve the image quality from metal artifacted data, our algorithm uses edge preserving interpolation method with BM3D smoothing algorithm. For evaluation, we used both real human dental data with metal artifacts and metal implemented phantom. Methods: Real metal artifacted data was obtained at Ewha womans university Mokdong hospital. Metal implemented phantom containing air, water, teflon, metalis made to evaluate the performance of the proposed algorithm. With obtained projection data, metal part is extracted from the back‐projection image. Once the metal location and size is obtained, all corrupted non‐metal part is recovered using edge preserving wavelet interpolation. With interpolated data, reconstruct image with filtered back‐projection. Apply BM3D to smooth the reconstructed image with edge preserving. BM3D is a state‐of‐art smoothing algorithm, which uses block information to get the similarity of the image. Replace the metal data from the smoothed image and continue doing this forward‐and backward projection iteratively. Our algorithm is based on the iterative reconstruction algorithm combined with the wavelet based interpolation with edge preserving smoothing filter. Results: We compared the proposed algorithm with Filtered back‐projection and iterative reconstruction algorithm with bilinear interpolation and Gaussian smoothing filter. For the evaluation, we used visual comparison, and quantitative comparison using student t‐test. For the real data, there is no exact solution. Thus only visual comparison can be applied. For the phantom data, we did the quantitative error comparison as well as visual comparison. The proposed algorithm outperformed results in both real and phantom data. Conclusion: We have developed a new CT reconstruction algorithm for metal artifact reduction and demonstrated its superiority compared with other algorithms.

Gradient Based Edge Preserving Interpolation and Its Application to Super-Resolution

Super-resolution is a process for obtaining high quality and high resolution images from a set of or only one low-resolution image. The most practical one for image super-resolution is reconstruction-based method, which minimizes the difference between observed low-resolution images and the estimation for high resolution image. Therein, interpolation step plays a key role for the estimated high resolution image quality. Usually, the conventional bilinear or bicubic methods are used in the reconstruction-based super-resolution. However, these conventional interpolations generally lead to blurring on edge regions and need more time for convergence in reconstruction-based super-resolution method. Therefore, this paper propose a gradient based edge preserving interpolation method, which can reduce not only artifact noise but also blurring near edge regions in the estimated high resolution image. Furthermore, our proposed interpolation method can also solve large complexity and time-consuming problem in the recently developed New Edge-directed interpolation, which usually can achieve sharp edge in the high resolution reconstructed image. Experiments validate that our proposed interpolation method for image super-resolution is more effective than the conventional interpolation ones.

A novel image interpolation method using the bilateral filter

In general, since the noise deteriorates the interpolation performance in digital images, it is effective to employ denoising prior to the interpolation. In this paper, we propose a novel interpolation framework in which denoising and image sharpening methods are embedded. In the proposed framework, the image is first decomposed using the bilateral filter into the detail and base layers which represent the small and large scale features, respectively. The detail layer is adaptively smoothed to suppress the noise before interpolation and an edge-preserving interpolation method is applied to both layers. Finally, the high resolution image is obtained by combining the base and detail layers. Experimental results show that the proposed algorithm outperforms the conventional methods while suppressing blurring and jagging.