In this paper, we present an efficient region-based image retrieval method, which uses multi-features color, texture and edge descriptors. In contrast to recent image retrieval methods, which use discrete wavelet transform (DWT), we propose using shape adaptive discrete wavelet transform (SA-DWT). The advantage of this method is that the number of coefficients after transformation is identical to the number of pixels in the original region. Since image data is often stored in compressed formats: JPEG 2000, MPEG 4…; constructing image histograms directly in the compressed domain, allows accelerating the retrieval operation time, and reducing computing complexities. Moreover, SA-DWT represents the best way to exploit the coefficients characteristics, and properties such as the correlation. Characterizing image regions without any conversion or modification is first addressed. Using edge descriptor to complement image region characterizing is then introduced. Experimental results show that the proposed method outperforms content based image retrieval methods and recent region based image retrieval methods.

In this paper, we present an approach for scene text extraction from natural scene video frames. We assumed that the planar surface contains text information in the natural scene, based on this assumption, we detect planar surface within the disparity map obtained from a pair of video frames using stereo vision technique. It is followed by extraction of planar surface using Markov Random Field (MRF) with Graph cuts algorithm where planar surface is segmented from other regions. The text information is extracted from reduced reference i.e. extracted planar surface through filtering using Fourier-Laplacian algorithm. The experiments are carried out using our dataset and the experimental results indicate outstanding improvement in areas with complex background where conventional methods fail.

Real-time object tracking is one of the most crucial tasks in the field of computer vision. Many different approaches have been proposed and implemented to track an object in a video sequence. One possible way is to use mean shift algorithm which is considered to be the simplest and satisfactorily efficient method to track objects despite few drawbacks. This paper proposes a different approach to solving two typical issues existing in tracking algorithms like mean shift: (1) adaptively estimating the scale of the object and (2) handling occlusions. The log likelihood function is used to extract object pixels and estimate the scale of the object. The Extreme learning machine is applied to train the radial basis function neural network to search for the object in case of occlusion or local convergence of mean shift. The experimental results show that the proposed algorithm can handle occlusion and estimate object scale effectively with less computational load making it suitable for real-time implementation.

Steganography is the art or science that is used in secret communication. It means that there is a secret message that is hidden within another cover media. The cover media may be image, video or audio and the secret message may be any type of digital message. The hidden message doesn't have any relationship with the cover media where the cover media is just to protect the secret message from hacking by unauthorized receiver. The audio cover is used in this paper because of the higher sensitivity of the human auditory system (HAS) than the human visual system (HVS). In this paper, we proposed a hybrid technique to audio steganography. This technique is based on a hybrid between two techniques of audio steganography. These techniques are Least Significant Bit (LSB) technique and modification of phase coding. The hybrid between them is for improving the performance of the phase coding where the performance of it is very low. Audio steganography performance is measured by several factors, the most important one of them is Signal to noise ratio (SNR) which is used to compare the performance of our technique with some known techniques.

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).

The security of digital image watermarking is improved by scrambling the watermark using different chaotic maps or cellular automata in such a way that an unauthorized person can't recover the watermark without the secret keys. In this proposed scheme three secret keys are used in which one key is used to make the watermark chaotic and other two keys are used for scrambling the cover image. In this scheme the cover image is scrambled by using the game of life cellular automation and the watermark is made chaotic by performing the X-OR operation between the binary watermark and logistic map. Although it increases the computational complexities, but the security of watermarking is improved by involving three secret keys. In addition, for ensuring imperceptibility and making the watermarking robust, a mask of size 3×3 is run over the scrambled cover image in which one bit of chaotic watermark is embedded in 3×3 block of cover image by modifying one of the neighbor pixels. Then the scrambled modified cover image is descrambled using game of life cellular automation for obtaining watermarked image. This proposed combined chaotic and cellular automata based watermarking scheme is compared with existing chaotic based watermarking schemes and gives satisfactory values of Peak Signal to Noise (PSNR), Mean Squared Error (MSE) and Normalized Correlation (NC).

As a contribution from research conducted by many, various image compression techniques have been developed on the basis of transformation or decomposition algorithms. The compressibility of a signal is seen to be affected by the entropy in the signal. Compressibility is high if the energy distribution is concentrated in fewer coefficients. It is reasonable to expect that sparse signals have a highly compressible nature. Thus, sparse representations have potential uses in image compression techniques. There are many techniques used for this purpose. As an alternative to these traditional approaches, the use of Discrete Rajan Transform for sparsification and image compression was explored in this paper. The simulation results show that higher quality compression can be achieved for images using Discrete Rajan Transform in comparison with other popular transforms like Discrete Cosine Transform, and Discrete Wavelet Transform. The results of the experiment were analyzed on the basis of seven quality measurement parameters – Mean Squared Error, Peak Signal to Noise ratio, Normalized Cross-Correlation, Average Difference, Structural Content, Maximum Difference, and Normalized Absolute Error. It was observed that Discrete Rajan Transform is effective in introducing sparsity in images and thereby improving compressibility.

Content-based image retrieval (CBIR) is broadly applicable for searching digital images from a gigantic database. Images are retrieved by their primitive visual contents such as color, texture, shape, and spatial layout. The approach presented in this paper utilizes structural connections within an image by integrating textured color descriptors and structure descriptors to retrieve semantically significant images. The retrieval results were obtained by applying the HSV histogram, color coherence vector, and local binary pattern histogram to the standard database of Wang et al., which has 1000 images of 10 different semantic categories. Euclidean distance was used to find the similarity between the query image and database images. This method was evaluated against different methods based on edge histogram descriptors, color structure descriptors, color moments, the color histogram, the HSV histogram, Tamura features, edge descriptors, geometrical shape attributes, and statistical properties such as mean, variance, skewness, and kurtosis. Retrieval results obtained using the proposed methods demonstrated a significant improvement in the average precision (73.8% and 73.1%) compared with those obtained using other existing retrieval methods.