Search

This paper presents a new wavelet-based multi-focus image fusion approach using method noise and anisotropic diffusion for two separate cases, i.e., with and without a reference image. It is specifically designed for real-time surveillance applications. It is a multi-step image fusion approach. Firstly, stationary wavelet transform (SWT) is performed to get low and high-frequency coefficients. Secondly, the input images' LL bands are fused using average operation. The rest of the respective bands are fused using a new correlation coefficient (CC) based fusion strategy using the threshold value calculated by structural similarity index metric (SSIM). Then inverse SWT is performed to reconstruct the fused coefficients. Thirdly, anisotropic diffusion-based method noise thresholding is introduced to recover the unprocessed and still damaged input images' components. Finally, the proposed approach's performance has experimented with various qualitative (visual perception) and quantitative factors (performance metrics). The experimental outcomes show that the proposed approach generates fine edges, high visual quality, high clarity of objects, and less degradation. The proposed multi-step hybrid technique is implemented to generate high-quality fused images. The experimental outcomes verify the achievement of the proposed approach.

This article discusses a hybrid wavelet-based multi- focus image fusion technique using Principal Component Analysis (PCA). Previously many authors have used many wavelet transforms in a different way for effective image fusion purpose but in this article, the authors have used Discrete Wavelet Transform (DWT) and Stationary Wavelet Transform (SWT) to process the input images in parallel. Later their results are further merged/ fused using standard PCA method. The results of the proposed method are compared with some of the conventional and non-conventional works related to SWT, DWT, and PCA. The result shows a good result than the compared method’s result. The results are analyzed and assessed using visual appearance of the final fused image and also using some of the parameters like Standard Deviation (SD), Petrovics metric (PM), Average Pixel Intensity (API) and Average Gradient (AG).

Image Fusion is the methodology to combining multiple images resulting in a single image keeping all the necessary aspects and features of original image with extended information. In this paper, we perform the comparative study between Stationary Wavelet Transform (SWT) and Discrete Wavelet Transform (DWT). We will discover various sectors where SWT performs better than DWT. Stationary Wavelet Transform (SWT) is primarily applied on the original image for obtaining the edge information of image, in Level 1 and Level 2 respectively. Next, both edge images are fused to get the finalised absolute edge image using spatial frequency measurement. Therefore, simple fusion techniques are used for comparing images.

Today’s research era, image fusion is a actual step by step procedure to develop the visualization of any image. It integrates the essential features of more than a couple of images into a individual fused image without taking any artifacts. Multi-focus image fusion has a vital key factor in fusion process where it aims to increase the depth of field using extracting focused part from different multiple focused images. In this paper multi-focus image fusion algorithm is proposed where non local mean technique is used in stationary wavelet transform (SWT) to get the sharp and smooth image. Non-local mean function analyses the pixels belonging to the blurring part and improves the image quality. The proposed work is compared with some existing methods. The results are analyzed visually as well as using performance metrics.

In digital science, organizations have to need secure transmission especially for image data. Images are encrypted to provide the secure transmission over the network. To encrypt, image processing concepts play a very important role. With this motivation, this paper is proposed image encryption in discrete orthonormal Stockwell transform (DOST) and CNN. For image encryption, pixel shuffling method and Arnold transform are used in DOST domain. The proposed methods are compared and analyzed the performance using mean error, PSNR and Entropy difference. The obtained results are better in compared to existing methods.

Today, Computed tomography (CT) is one of the high efficient tools in medical science which helps to diagnose the human body. The presence of noise may degrade the visual quality of the CT images, especially for low contrast images. Therefore, we propose a method based on the modification of total variation (TV) for noise suppression in CT images which is helpful to preserve the clinically relevant details. The modification of TV is performed by introducing a new exponentially directional weighted function (EDWF), which is based on the difference between L1 and L2 norms over the exponential function. Furthermore, a numerical algorithm is designed to solve the minimization problem of EDWF using Split Bregman method. The experimental results of proposed scheme are visually analyzed over the real noise CT image, added noise in true CT images, and also on low contrast zoomed noisy CT image. Apart from visual analysis, the proposed scheme is also verified with some standard performance metrics (RMSE, PSNR, SSIM, ED, DIV and GMSD). The proposed scheme is also compared with some standard existing methods and it is observed that performance of proposed scheme is superior to existing methods in terms of visual quality and performance metrics. Keyword: Total variation, Computed tomography, Image denoising, Anisotropic function, Isotropic function

In cloud computing environment, various image processing approaches have been processed for different applications. Image processing facility in cloud computing is broadly used in numerous fields such as medical science, satellite applications, security applications and so on. The main focus of image processing in cloud computing environment is that the cloud service as image processing services is required with higher quality of image processing. Therefore, cloud based image processing services are also required. Hence, this paper shows some of the recent image processing applications in the cloud environment which are reviewed and discussed. Some popular image processing tools which are working in the cloud environment are also reviewed.

The impact of radiation dose is directly related to the quality of computed tomography (CT) images. Low-dose CT images are degraded with the noise and other factors. Noise reduction methods are very helpful to enhance the noisy CT images with a possibility to increase the signal-to-noise ratio (SNR) and have a scope for further reduction of radiation dose. In this study, a denoising scheme is proposed which is applicable only for two identical images with uncorrelated noise. In the proposed scheme, a non-local means (NLM) filter is used to denoise the first input image and a wavelet packet thresholding to denoise the second input image. Results of NLM filter are analysed and found excellent for noise suppression but missing the small structures of the input image. To recover that, the proposed scheme is using correlation-based wavelet packet thresholding. The final outcomes of the proposed scheme are excellent in terms of noise suppression and structure preservation. The proposed scheme is compared with existing methods and it is observed that performance of the proposed method is superior to existing methods in terms of visual quality, image quality index, peak SNR and entropy difference.

CT imaging is widely used in medical science over the last decades. The process of CT image reconstruction depends on many physical measurements such as radiation dose, software/hardware. Due to statistical uncertainty in all physical measurements in Computed Tomography, the inevitable noise is introduced in CT images. Therefore, edge-preserving denoising methods are required to enhance the quality of CT images. However, there is a tradeoff between noise reduction and the preservation of actual medical relevant contents. Reducing the noise without losing the important features of the image such as edges, corners and other sharp structures, is a challenging task. Nevertheless, various techniques have been presented to suppress the noise from the CT scanned images. Each technique has their own assumptions, merits and limitations. This paper contains a survey of some significant work in the area of CT image denoising. Often, researchers face difficulty to understand the noise in CT images and also to select an appropriate denoising method that is specific to their purpose. Hence, a brief introduction about CT imaging, the characteristics of noise in CT images and the popular methods of CT image denoising are presented here. The merits and drawbacks of CT image denoising methods are also discussed.

In Computed Tomography (CT), image degradation such as noise and detail blurring is one of the universal problems due to hardware restrictions. The problem of noise in CT images can be solved by image denoising. The main aim of image denoising is to reduce the noise as well as preserve the important features such as edges, corners, textures and sharp structures. Due to the large capability of noise suppression in noisy signals according to neighborhood pixels or coefficients, this paper presents a new technique to denoise CT images with edge preservation in tetrolet domain (Haar-type wavelet transform) where a locally adaptive shrinkage rule is performed on high frequency tetrolet coefficients in such a way that noise can be reduced more effectively. The experimental results of the proposed scheme are excellent in terms of noise suppression and structure preservation. The proposed scheme is compared with some standard existing methods where it is observed that performance of the proposed scheme is superior to the existing methods in terms of visual quality, MSE, PSNR and Image Quality Index (IQI).

In today's era, image fusion is the major technique to enhance the images using multiple images. In this paper, cross-bilateral based image fusion is proposed in wavelet domain. Firstly, images are decomposed using wavelet transform. Approximation parts are fused using weighted average function while detail part is processed by max operation. First intermediate result is obtained using inverse DWT. Original input image is subtracted by first intermediate result. Over subtracted image, cross-bilateral filter is applied to get second intermediate result. Both first and second intermediate results are added to get the final result. The results are tested by various performance metrics and it was found that proposed method gives better outcomes in compare to existing methods.

In medical science, the role of medical images is very crucial to identify the diagnosis. Variety of medical images such as CT (Computed Tomography) images, MRI (magnetic resonance imaging) images are available but many times these different modalities need more informatics images. Therefore, multi-modality medical image fusion technique is one of the solutions which can provide more informatics details by fusing multi-modality medical images. In this paper, a new approach of multi-modality medical image fusion approach is proposed where fuzzy local information based C-means clustering and local entropy fusion approach is performed in SWT (stationary wavelet transform) Domain. To measure the performance of proposed scheme, a comparative study is also performed over the existing methods. The result analysis is performed in both visually and also by performance metrics such as mutual information, edge based similarity and blind image quality. From experimental evaluation, it can be analyzed that proposed scheme gives better outcomes in compare to existing methods.

The digital image carries a lot of pictorial information which has now become one of the chief ways of communication in this generation. The transmission of media especially image is often corrupted by the Gaussian noise due to various issues. This noise is needed to be removed. Image denoising technique is applied to remove such noise and make it of high quality. This additive Gaussian noise can be removed using wavelet denoising technique. This paper presents a better method that controls the threshold (T) adaptively to eliminate noise. Experimental results of the proposed method are better than the existing denoising algorithm Such as NSTISM(19), SPBIDM(18), BayesShrink, NormalShrink, and ModifiedBayesShrink.

In modern science there is a rapid development of artificial intelligence, image processing has gradually fascinated and inspired the attention of many researchers in the field of artificial intelligence and has become an interesting and demanding task. The main idea of Image caption is to automatically generate natural language descriptions according to the information observed in an image, this is an important portion of scene understanding, which combines all the knowledge and information available of computer vision and natural language processing. The use of image caption is broad and noteworthy, for example, the understanding of human-computer collaboration. This paper reviews the related methods and focuses on the attention mechanism, which plays a vital role in computer vision and is broadly used in image caption generation tasks. Furthermore, the advantages and the shortcomings of these methods are discussed, providing the commonly used datasets and evaluation criteria in this field. Finally, this paper proposes some open challenges in the image caption task.

Most of the local descriptors exists today are based on the relationship between the neighborhood pixels and the center pixel by considering the 3 × 3 spatial window. Moving one step forward from the previous methodologies used, this research paper introduces the novel descriptor for Face analysis called as orthogonal difference-Local binary pattern (OD-LBP). In OD-LBP, initially the 3 gray level differences are computed for each orthogonal position (of 2 groups, in 3 × 3 pixel window) by subtracting the 2 closest neighborhoods and the center pixel from the respective orthogonal value. Then these respective gray level differences are compared with the value produced after employing the novel comparison method. Finally the 1 × 24 size row vector is produced by concatenating the binary patterns produced from both the orthogonal groups. Eventually the 24 bit row vector is split into three 8 bit binary patterns from which the 3 OD-LBP codes are produced. By employing this concept for each pixel position eventually results in the 3 OD-LBP transformed images. All the 3 transformed images are then divided into 3 × 3 sub-regions for histogram extraction. The combined sub-regional histograms from all the transformed images are the entire feature size of the OD-LBP descriptor. To achieve the compact feature representation PCA is applied further and then classification is performed by the SVMs. The entire concept is tested on 5 challenging databases which include ORL, GT, JAFFE, MIT-CBCL and Yale. The proposed descriptor (i.e. OD-LBP) achieves very promising results on all 5 databases as it comprehensively outperforms the other 10 state of art descriptors LBP, CS-LBP, 6 × 6 MB-LBP, OC-LBP, MBP, HOG, ELBP, OS-LTP, WCLBP and RLBP on majority of sub-sets. Numerous states of approaches from the literature are also outclassed by proposed descriptor.

This research paper introduces two novel color descriptors for Face Recognition (FR) called as Color ZigZag Binary Pattern (CZZBP) and Color Median Block ZigZag Binary Pattern (CMBZZBP). The pixels aligned in ZigZag manner are utilized for producing the feature size of both the proposed descriptors. For first proposed descriptor, the ZigZag features are extracted from 3 color components of the RGB color space format, which are concatenated to produce the entire feature size of the CZZBP descriptor. Specifically, on 3 color components i.e. R, G and B, the features are extracted by utilizing three different type of ZigZag binary patterns called as ZZBP1, ZZBP2 and ZZBP3. The ZZBP1 is applied on the R component, ZZBP2 is applied on the G component and ZZBP3 is applied on the B component. In the second proposed descriptor, the same concept is employed on medians computed from 9 different blocks of 9 × 9 pixel window, where each block size is 3 × 3. To produce compact feature size, PCA is employed and classification is performed by SVMs and NN. The 2 databases used for performance evaluation are GT and Faces94. Experiments reveals effectiveness of the proposed descriptors than compared descriptors. Among all descriptors, CMBZZBP is most impressive.

In medical science, Computed Tomography is one of the powerful tools to reconstruct 3D image with measuring stack of parallel slices, where Radon transform is useful to get image slices by set of line integrals. 3D reconstructed images are very helpful to analyze multiple abnormalities in comparison of 2D images. Only CT scanned images is not sufficient to get the exact position and surface information of patients. This paper is dealing with the problem of direct 3D-reconstruction of medical images. We examined 3D surface reconstruction over 2D-CT scanned image and also tested over the simulated image. Shape from shading (SFS) technique is useful to get the geometric information for recover features from variation of shading. In this paper, Fast Marching Method (FMM) is used to get the 3D surface over the medical images (i.e., real image and synthetic image).

Multi-focus image fusion is an essential method to merge the important information from a data-set of images with the similar scene. The fused image can be more informative than any previous source image. In this paper, a new scheme based on discrete wavelet transformation (DWT) is proposed. Multi-focus images are decomposed in different level and get the relevant results with few fusion metrics like standard deviation (SD), coefficient correlation (CC), entropy etc. In between, variance based fusion on approximation part and weighted fusion on detailed part have been used to generate the further process. Finally multi-focus images are all in one focus. The experimental work is applied to produce the good quality scale images; researched work established the qualitative approach impressively.

Noise in CT images is very common problem which degrade the quality of CT images. In this paper, a method is proposed where CT images are denoised using bilateral method with the concept of Bayes Shrinkage rule in wavelet domain. Firstly, noisy CT image is filtered using bilateral filter. This filtered image is suppressed by noisy input image. Over the subtracted image, wavelet packet based thresholding is performed. To get the final denoised image, the thresholded output image is added with bilateral filtered image. To analyze the efficiency of propose algorithm, a comparative result analysis has been performed with some recent similar methods and also from some state of arts. The comparative analysis indicates that in most of the cases, the proposed algorithm gives better results.

The unwanted data obtained through the medical image fusion is the main problem in biomedical applications, guided-image surgical and radiology. The Stationary Wavelet Transform (SWT) denoted the various advantages over conventional representation of imaging approach. In this research article we introduced innovative multi-modality fusion technique for medical image fusion based upon the SWT. In our approach it disintegrates of source images into approximation layers (coarse layer) and detail layers through the Stationary Wavelet Transform scheme, then applying of the Fuzzy Local Information C-means Clustering (FLICM) and Local contrast fusion approach to overcome the blurring effect, sensitiveness and conserve the quality evaluation in the distinguish layers. The demonstration shows that it preserves more detailed information in the source images and it enhances the more quality features and edge preserved of the final fused image obtained through the reconstruction procedure by recursive initial steps. The different methodologies with other techniques to evaluate performance such as mutual information, edge based similarity and blind image quality. This shows that in both the objective and subjective analysis our methodology results attained more supercilious performance.

The main aim of image denoising is to improve the visual quality in terms of edges and textures of images. In Computed Tomography (CT), images are generated with a combination of hardware, software and radiation dose. Generally, CT images are noisy due to hardware/software fault or mathematical computation error or low radiation dose. The analysis and extraction of medical relevant information from noisy CT images are challenging tasks for diagnosing problems. This paper presents a novel edge preserving image denoising technique based on wavelet transform. The proposed scheme is divided into two phases. In first phase, input CT image is separately denoised using different patch size where denoising is performed based on thresholding and its method noise thresholding. The outcome of first phase provides more than one denoised images. In second phase, block wise variation based aggregation is performed in wavelet domain. The final outcomes of proposed scheme are excellent in terms of noise suppression and structure preservation. The proposed scheme is compared with existing methods and it is observed that performance of proposed method is superior to existing methods in terms of visual quality, PSNR and Image Quality Index (IQI).

Recently in medical imaging, various cases of cancers have been explored because of high dose radiation in computed tomography (CT) scan examinations. These high radiation doses are given to patients to achieve good quality CT images. Instead of increasing radiation dose, an alternate method is required to get high quality images for diagnosis purpose. In this paper, we propose a method where, the noise of CT images will be estimated using patch-based gradient approximation. Further, estimated noise is used to denoise the CT images in tetrolet domain. In proposed scheme, a locally adaptive-based thresholding in tetrolet domain and non-local means filtering have been performed to suppress noise from CT images. Estimation noise from proposed method has been compared from added noise in CT images and it was observed that noise is almost correctly estimated by proposed method. To verify the strength of noise suppression in proposed scheme, comparison with recent other existing methods have been performed. The PSNR and visual quality of experimental results indicate that the proposed scheme gives excellent outcomes in compare to existing schemes.

Edge detection is a fundamental tool in image processing. Several edge detectors have been proposed in literature for enhancing and detecting edges in images. Image Edge detection significantly reduces the amount of data and filters out useless information, while preserving the important structural properties in an image. In this paper, the application of two-dimensional cellular automata using Moore Neighborhood has been proposed for edge detection. The idea is simple but effective technique for edge detection. Edge basically occurs where there is significant change in intensity. The principle of the algorithm used is to increase the difference between those pixels where intensity values change significantly. So by using this concept, detected edges are wider and clear. The given algorithm can be applied to gray scale and monochrome images.

Biometrics authentication system promises to deliver a level of uniqueness to identification applications that uses Tongue print, fingerprint or facial scans and iris or voice recognition to identify users. The tongue is a unique organ in that it can be stuck out of mouth for inspection, in this act offering a proof of life, and yet it is otherwise well protected in the mouth and is difficult to forge. The tongue also presents both geometric shape information and physiological texture information which are potentially useful in identity verification applications. Tongue biometric can function as an extremely reliable means for personal identification and act as a general biometric in all applications. This paper demonstrates two phases. First, find out spots on the tongue with the help of histogram and second phase to extract image tongue and recognize from the tongue-image database.

Computed Tomography (CT) is one of the most widespread radiological tools in medical diagnostics. To achieve good quality of CT images with low radiation dose has drawn a lot of attention to researchers. Hence, post processing of CT images has become a major concern in medical image processing. This paper presents a novel edge-preserving image denoising scheme based on Dual-tree Complex Wavelet Transform (DT-CWT), Bilateral filtering and a locally adaptive thresholding method. The noisy image is decomposed into Complex Wavelet coefficients through a Dual-tree Complex Wavelet Transform. Low-pass subbands are modified using Bilateral filter. High pass subbands are modified using locally adaptive thresholding based on interscale statistical dependency, where the noise variance of noisy wavelet coefficients are estimated using a robust median estimator. Denoised image is retrieved using inverse DT-CWT. The proposed scheme is compared with existing methods. It is observed that performance of proposed method is superior than existing methods in terms of visual quality, Image Quality Index (IQI) and Peak Signal-to-noise Ratio (PSNR).

Computed Tomography (CT) is streamlined in radiological diagnostics and has become an imperative tool in medical examinations. The difficulty that arises with the demand is to improve CT image quality without increasing dose. In this paper, Wavelet based noise reduction technique is proposed to improve image quality where adaptive Wiener filtering and Wavelet Packet Threshold (WPT) algorithm are applied. The Noisy CT image is decomposed using DWT, where approximation part is filtered using WPT algorithm and detail part is filtered by the adaptive Wiener filtering. By using the level dependent, the wavelet packet tree coefficients are calculated using optimal linear interpolation shrinkage function. Denoised image is acquired using wavelet packet reconstruction and inverse DWT. The value of the peak signal to noise ratio (PSNR) is used as the measure of image visual quality. Experimental results demonstrate that the proposed method improves the image visual quality in respect of noise removal and edge preservation.

Object tracking is an important task within the field of computer vision. There are two key steps in video analysis: detection of interesting moving objects and tracking of such objects from frame to frame. Mean shift object tracking algorithm is a feature based algorithm. In this paper color information of object are extracted as well as texture information is also extracted by using local binary pattern technique to represent the object. In joint color-texture histogram region of interest is extracted and then edges and corners are extracted from the region of interest. Color histogram is insensitive to nondeformation and partial occlusion but it does not perform well when in case of color aberration. To overcome this drawback Local Binary Pattern method is used. This feature makes this algorithm more efficient in comparison of color histogram algorithm.