Image interpolation for progressive transmission by using radial basis function networks

New Efficient Interpolation Techniques for Medical Images.. [DOI: 10.21203/rs.3.rs-2869416/v1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0]

This paper presents new efficient six image interpolation techniques that aim to obtain high resolution images from low resolution ones. The problem of dealing with image interpolation as an inverse problem was treated in wavelet domain through the first proposed approach. This approach estimates wavelet coefficients in high frequency sub-images using least squares algorithm. The approach reveals performance improvement in PSNR and MSE values compared to bilinear and least squares algorithm. A second approach based on using artificial neural network for modeling the maximum entropy interpolation algorithm was presented. Training the suggested network is achieved by adjusting its weights using learning algorithm to minimize the sum of squared error between target and actual output of output neurons. The proposed approach requires fewer computations than the conventional high-quality image interpolation methods with little reduction in PSNR compared to the maximum entropy algorithm. A third algorithm which combines new contourlet transform and edge-based interpolation algorithm was proposed to improve object boundaries regularity which could be achieved by treating each successive approximation of the high-resolution image as a noisy approximation, hence enforcing the contourlet coefficients to have a sparsity constraint. The proposed algorithm gives an improvement of 2 dB in PSNR compared to IIGR algorithm and achieves the best MSE and correlation coefficient. The fourth approach presents an efficient method to get HR images from the fusion of Magnetic resonance and computed tomography images by adopting least-squares strategy to iteratively obtain wavelet sub-bands of the target HR image. The proposed approach was found to achieve good results especially when the curvelet transform is used to merge the CT and MR images. The fifth approach presents neural modeling of the maximum entropy image interpolation algorithm based on both curvelet and wavelet fusion. Through this approach the fusion process is performed while edges with a curved nature are preserved. Results showed that applying curvelet transform in the fusion of CT and MR images is better than DWT fusion and that interpolating the fused image by neural modeling of maximum entropy image interpolation algorithm is superior to interpolating the original MR and CT images. The sixth approach presents a new image scale-up, super-resolution, approach based on image fusion principle. MR and CT images were scaled-up using sparse-representation modeling with dictionary learning algorithms of Yang et. al. and Michael et. al.. The images were fused by discrete wavelet and curvelet transforms, then scaled-up by the same algorithms. Simulation results showed that scaling-up the fused MR and CT images, either by wavelet or curvelet fusion techniques, gives higher PSNR values than scaling-up the MR and CT image separately.

A unified learning framework for single image super-resolution

It has been widely acknowledged that learning- and reconstruction-based super-resolution (SR) methods are effective to generate a high-resolution (HR) image from a single low-resolution (LR) input. However, learning-based methods are prone to introduce unexpected details into resultant HR images. Although reconstruction-based methods do not generate obvious artifacts, they tend to blur fine details and end up with unnatural results. In this paper, we propose a new SR framework that seamlessly integrates learning- and reconstruction-based methods for single image SR to: 1) avoid unexpected artifacts introduced by learning-based SR and 2) restore the missing high-frequency details smoothed by reconstruction-based SR. This integrated framework learns a single dictionary from the LR input instead of from external images to hallucinate details, embeds nonlocal means filter in the reconstruction-based SR to enhance edges and suppress artifacts, and gradually magnifies the LR input to the desired high-quality SR result. We demonstrate both visually and quantitatively that the proposed framework produces better results than previous methods from the literature.

An adaptable time-delay neural-network algorithm for image sequence analysis

In this letter we present an algorithm based on a time-delay neural network with spatio-temporal receptive fields and adaptable time delays for image sequence analysis. Our main result is that tedious manual adaptation of the temporal size of the receptive fields can be avoided by employing a novel method to adapt the corresponding time delay and related network structure parameters during the training process.

Progressive cross-section display of 3D medical images

The paper presents a hierarchical coding algorithm for 3D medical images based upon hierarchical interpolation with radial basis function networks. By using the properties of the Kronecker product, the computation of the network parameters and the 3D image reconstruction are efficiently done in (L4) computation time and O(L3) storage space, when applied to 3D images of size (L x L x L). A further reduction in processing time is accomplished by using sparse matrix techniques. The salient features of the proposed coding method are that arbitrary cross-section images can be progressively displayed without reconstruction of the whole 3D image; the first image reconstruction starts as soon as the first data transmission has been completed; no expanding procedure is required in 3D image reconstruction, and the blocking effects are not apparent even in the lowest-resolution image. Experimental results using two 3D MRI images, of size (128 x 18 x 64) and with 8-bit grey levels, show that the coding performance is better than that of the 3D DCT coding by about 0.25 bits pixel-1 at higher bit rates, and that the new cross-section display method synthesises the coarsest (finest) section image about six (three) times faster than the standard method that requires the whole 3D image reconstruction.