A three-dimensional (3-D) segmentation technique is proposed, which separates the 3-D reconstructed volume data into several sub-regions having similar characteristics. This method is based on iteratively merging and splitting the sub-regions in the space of diffusion tensor fields. The superquadrics have been used as a figure of merit to measure the similarity between the neighboring voxels. By using the segmented 3-D in the magnetic resonance angiography data, we can efficiently visualize the vascular structure without any prior information as well as improve the medical diagnosis and therapy.

Through-the-wall radar imaging (TWRI) is emerging as a viable technology for providing high-quality imagery of enclosed structures. TWRI makes use of electromagnetic waves to penetrate through building wall materials. Due to the “see” through ability, TWRI has attracted much attention in the last decade and has found a variety of important civilian and military applications. Signal processing algorithms have been devised to allow proper imaging and image recovery in the presence of high clutter, which is caused by front walls and multipath due to reflections from internal walls. Recently, research efforts have shifted toward effective and reliable imaging under constraints on aperture size, frequency, and acquisition time. In this respect, scene reconstructions are being pursued with reduced data volume and within the emerging compressive sensing (CS) framework. We present a review of the CS-based scene reconstruction techniques that address the unique challenges associated with fast and efficient imaging in urban operations. Specifically, we focus on ground-based imaging systems for indoor targets. We discuss CS-based wall mitigation, multipath exploitation, and change detection for imaging of stationary and moving targets inside enclosed structures.

Different camera configurations to capture panoramic images and videos are commercially available today. However, capturing omnistereoscopic snapshots and videos of dynamic scenes is still an open problem. Several methods to produce stereoscopic panoramas have been proposed in the last decade, some of which were conceived in the realm of robot navigation and three-dimensional (3-D) structure acquisition. Even though some of these methods can estimate omnidirectional depth in real time, they were not conceived to render panoramic images for binocular human viewing. Alternatively, sequential acquisition methods, such as rotating image sensors, can produce remarkable stereoscopic panoramas, but they are unable to capture real-time events. Hence, there is a need for a panoramic camera to enable the consistent and correct stereoscopic rendering of the scene in every direction. Potential uses for a stereo panoramic camera with such characteristics are free-viewpoint 3-D TV and image-based stereoscopic telepresence, among others. A comparative study of the different cameras and methods to create stereoscopic panoramas of a scene, highlighting those that can be used for the real-time acquisition of imagery and video, is presented.

High-dynamic-range (HDR) imaging has become more and more popular in recent years, and a number of methods for the realistic representation of natural scenes have been proposed. Gradient domain HDR compression is one of the most popular approaches; however, their results are not satisfactory due to the parameter setting. Motivated by perceptual studies on image quality, we propose a novel HDR compression method to achieve better image quality. In the proposed method, we combine global and local tone-mapping operators, and manipulate the HDR compression in both luminance and gradient domains to improve the appearance and naturalness of results. Moreover, we deal with the parameter setting problem effectively in HDR compression by automatic k factor decision. Extensive experiments on various test images demonstrate that the proposed method is very effective in enhancing the overall quality of the image as well as maintaining the details without the loss of naturalness.

A good superresolution (SR) algorithm obtains high-resolution (HR) images from the corresponding low-resolution (LR) ones and, moreover, makes the former look like they had been acquired with a sensor having the expected resolution or at least as “natural” as possible. In general, fast SR algorithms usually result in more ill artifacts in the enlarged image, while the well-performed ones usually have great complexity and take much more computing time. For this purpose, four efficient SR algorithms based on regression models are proposed. In the proposed SR algorithms, the difference of a natural HR image and an HR image obtained by fast interpolation is taken as the lost detail and is supposed to be composed of several different oriented details. By the self-similarity of the input LR image and its corresponding HR image, a regression model is established by the input LR image to decide the proper respective weights of these oriented details which is then used to reconstruct the lost detail of the natural HR image. As shown in the experimental results, the proposed SR algorithms not only perform well in both objective criteria and visual quality but also take less computing time than some well-performing algorithms.

A color-transfer method that can transfer colors of an image to another for the local regions using their dominant colors is proposed. In order to naturally transfer the colors and moods of a target image to a source image, we need to find the local regions of colors that need to be modified in the image. Since the dominant colors of each image can be used for the estimation of color regions, we develop a grid-based mode detection, which can efficiently estimate dominant colors of an image. Based on these dominant colors, our proposed method performs a consistent segmentation of source and target images by using the cost-volume filtering. Through the segmentation procedure, we can estimate complex color characteristics and transfer colors to the local regions of an image. For an intuitive and natural color transfer, region matching is also crucial. Therefore, we use the visually prominent colors in the image to meaningfully connect each segmented region in which modified color-transfer method is applied to balance the overall luminance of the final result. In the Experimental Result section, various convincing results are demonstrated through the proposed method.

We propose a new technology of optical watermarking that uses a one-dimensional high-frequency pattern to protect the portrait rights of three-dimensional (3-D) shaped real objects by preventing the use of images captured illegally with cameras. We conduct experiments using human faces as real 3-D objects assuming that this technology would be applied to human faces to protect their portrait rights. We utilize the phase difference between two color component patterns, i.e., binary information was expressed if the phase of the high-frequency pattern was the same or opposite. The experimental results demonstrate that this technique was robust against the pattern being deformed due to the curved surface of the 3-D shaped object, and the highest degree of accuracy of 100% in reading out the embedded data was possible by optimizing the conditions under which data were embedded. As a result, we can confirm that the technique we propose is feasible.

An important problem associated with particle filtering is track drift, which is caused by the inaccurate likelihood measurement function and state model; another problem is the heavy computation cost in high state space. A human body is an articulated object with high degrees of freedom, and humans often perform versatile unconstrained motions. Therefore, these two problems are invariably encountered when human motion is reconstructed from monocular video sequences using particle filters. To overcome these problems, we present a novel approach to recover three-dimensional human upper-body pose using a combination of a deterministic and stochastic method that takes full advantage of the benefits of both methods—highly accurate pose reconstruction and low computation cost. The reconstruction of the human upper-body pose is divided into two parts: global pose estimation and pose estimation of the remaining joints. The global pose is determined by solving a system of six nonlinear equations established by three scale invariant feature transform (SIFT) correspondences within the left and right shoulder segments. Estimation of the pose of the remaining joints is accomplished using two particle filters, one for left arm pose estimation and the other for right arm pose estimation. The image projection of the segment model is obtained by forward kinematics under a perspective camera model, and the likelihood measurement functions involving two features are presented to enhance the model fitting performance. To avoid track drift, the particle filters can reinitialize particle sets if most of the particles deviate from the target object. The probability distribution of the new particle set is modeled by a Gaussian mixture model, in which each Gaussian is determined by matched SIFT correspondences. Experimental results show that our proposed approach effectively and efficiently achieves human upper-body tracking and is more accurate than the standard particle filter.

Authentication of content in printed images poses a challenge that cannot be addressed by conventional digital signature schemes because under the analog transport provided by the printing channel the verifier does not have access to the original digital content in pristine form. We present a method for cryptography-based authentication of the content in printed images that also provides the capability for identifying localized changes made by informed malicious attackers—key functionality that is missing in print scan robust hashes that have traditionally been used for print content authentication. The proposed method operates by embedding, within the printed image, an authentication signature that consists of an encrypted thumbnail of the image using a high capacity data hiding method for halftone images. To authenticate the content, the embedded signature is extracted from a scan of the printed image and, after decryption, compared with the printed content. An implementation of the method that incorporates human or automated verification and identifies potential local tampering by informed malicious attackers is developed and successfully demonstrated.

The total variation-regularization-based scheme for image reconstruction contaminated by Poisson noise is investigated. To quickly obtain the best numerical solution, a fast alternating direction algorithm is addressed and the convergence of the proposed approach is strictly proved. Subsequently, our numerical simulations markedly demonstrate the high efficiency and stability of the new strategy in terms of evaluating the computation speed and especially the quality of the recovered images, compared with some of the current state-of-the-art numerical algorithms.

The steganalysis/detection of spatial domain least significant bit (LSB) matching steganography in grayscale images, which is the antetype of many sophisticated steganographic methods, is concentrated on. Spatial LSB matching can be modeled by adding independent noise to the image, and it is proved theoretically that the LSB matching will smooth the image histogram and histogram of difference image. Accordingly, the absolute differences between adjacent elements of image histogram are calculated as the histogram features, and co-occurrence matrix is utilized to extract features based on image correlation. A calibrated image is generated by embedding a message into the pending image. The features are extracted from both pending and calibrated images, and the ratios of corresponding features between pending and calibrated images are used as the final features. A support vector machine is utilized to train the classifier with the extracted features. Experimental results show that the proposed features outperform some previous ones and reveal the respective strong points of histogram and correlation features in the detection of never-compressed and JPEG-compressed images.

A robust image watermarking technique based on a modification of the luminance component of the host color image is described. Three methods are proposed in the watermarking scheme in order to improve its performance in terms of accuracy of the extracted watermark and robustness of the embedded watermark. These methods are a new approach to watermark embedding in the luminance component of a host image and a new original image prediction technique in the watermark extraction process. A set of experiments is carried out to verify our proposed methods. The experimental results show significant improvements gained from our proposed watermarking scheme compared to previous existing schemes. The results also show enhanced robustness of the embedded watermark against various types of attacks. Our proposed watermarking scheme can be used for both color and gray scale images.

A novel support vector machine (SVM)-based boundary recovery procedure for segmented medical objects in low-resolution images is proposed. The proposed procedure consists of two steps: segmentation and boundary interpolation steps. First, we initially estimate a coarse object region using an active contour-based segmentation method. Boundary recoveries from the first step exhibit considerably blocky artifacts and are easily misled by noise. Then, a reliable boundary recovery is achieved in the next step by the proposed support vector machines based interpolation scheme. In simulation, the proposed algorithm shows more reliable and better performance in the presence of noise and adequately preserves shapes and smooth boundaries that are essential characteristics of medical objects. We illustrate it using real-life data sets in regard to nonconvex tube detection in wall shear stress, lumen detection in carotid stenosis, micro-calcifications detection in digital mammography, and nonmedical fields as well.

Considering statistical characteristics of multiwavelet coefficient vectors, an image fusion method with a window-based salience measure for vector fusion is proposed in the multiwavelet domain. After analyzing joint probability distribution of multiwavelet coefficient vectors, we respect the joint probability distribution as a peaky and heavy tailed two-dimensional non-Gaussian distribution, and hence model it by using a bivariate symmetric alpha-stable (BiSαS) distribution. Model parameters of the BiSαS distribution can be estimated from samples of multiwavelet coefficient vectors. In a sliding window, the estimated parameters can reveal the impulsive shape of the histogram of the multiwavelet coefficient vectors within the window. So we employ these model parameters to form the new salience measure indicating the degree of prominent information. In our method, vector fusion rule is introduced to avoid inconsistency of decision maps between the component sub-bands and, consequently, to lessen ringing artifacts in final fusion results. Experimental results support the effectiveness of our proposed method in the area of visual quality and objective evaluations.

A better scheme is proposed for image authentication and restoration based on standard deviation (STD) using two watermarks, one for detecting tampered regions and the other for retrieval. The main contribution of this work is a decrease of recovery watermark, which is compressed from three categories according to the STD values of every quantization discrete cosine transformation coefficients block. In addition, blocks corresponding to different categories are encoded to variable lengths and then encrypted based on the index of a chaotic sequence in an ascending order. At the same time, the authentication watermark uses a pixel-wise scheme obtained by random permutations and suitable exclusive OR operations among bitplanes of the host image. Experimental results show that the proposed scheme is not only secure but also detect any modification made to images and restore regions that have been modified.

A new approach is proposed for edge detection in ultrasound. The technique examines the image intensity profile for unit roots based on the Dickey–Fuller statistical test of stationarity. The existence of the unit root is a sign of nonstationarity and a possible edge. A simple algorithm to build a segmentation method based on this edge detection approach is also proposed, which is capable of delineating the perimeter of hollow structures such as blood vessels and cysts. In this approach, the radial edge profiles originating from the center of the object of interest are scanned for the change from stationary to nonstationary status. The algorithm treats the radial intensity profiles as a time series and uses the Dickey–Fuller statistical test along the radii to find the location at which the profile becomes nonstationary. A priori criteria for edge continuity, shape, and size of the object of interest are applied to enhance the stability of the algorithm. The accuracy is demonstrated on simulated ultrasound. Further, the method is examined on two different image sets of blood vessels and validated based on contours marked by experts. The worst case distance from expert contours is 1.8±0.3  mm . The average area difference between the expert and the extracted contours is ∼6% and ∼4% of the vessel area in the two datasets. The proposed segmentation method is also compared to segmentation using active contours on ultrasound images of breast and ovarian cysts and shown to be accurate and stable.

Although state-of-the-art image denoising algorithms have achieved outstanding results, removing real, color noise from a single image remains a challenging problem. An adaptive image denoising algorithm that integrates a constant time bilateral filter with noise-level estimation is proposed. The estimation of the noise-level function (NLF), which describes the noise level as a function of image brightness, is the key to ensure the removal of the color noise. To achieve this aim, a bilateral median filter is exploited to estimate the upper bound of NLF by fitting a lower envelope to the standard deviations of per-segment image variances. Furthermore, we make an empirical study on the relationship between the optimal parameter of constant time bilateral filter and the noise level. Then, an adaptive denoising algorithm, where the filtering parameter is automatically adjusted according to the estimated noise level, is conducted to obtain the underlying clean image from the noisy input. In addition, we present a new method of synthesizing noise, where the synthetic noise is very close to the real noise. Meanwhile, we test our algorithm on the synthetic noise images and on the real applications as well. Various experimental results show that our algorithm outperforms state-of-the-art denoising algorithms in eliminating real, color noise.

Linear scan cone-beam computed tomography (LSCBCT) is a technically simple computed tomography (CT) configuration and is powerful enough to inspect long objects. As one kind of limited-angle problem, the image reconstruction of LSCBCT is ill posed. However, the total variation minimization (TVM)-projection on convex sets (POCS) reconstruction algorithm, which is based on the TVM and POCS, has been proven effective for solving the limited-angle problem. While applying the TVM-POCS algorithm to the LSCBCT reconstruction, the reconstructed image is distorted near the edges of the object. To solve this problem, an improved iterative reconstruction algorithm was developed. The improved method integrated simultaneous algebraic reconstruction technique, TVM, and C-V model. The C-V model can detect objects whose boundaries are not necessarily defined by gradient. The developed algorithm can reduce artifacts by piecewise constant modification and get a more accurate image. Numerical simulations are presented to illustrate the effectiveness of the algorithm. Moreover, the developed algorithm can be applied to other x-ray CT reconstruction problems.

We present a line scan color vision system that detects aspect flaws occurring on the color surfaces of drinking glasses decorated due to an industrial silk-screen process. For this purpose, we have designed a specific image acquisition device based on a color line-scan camera. As the pattern printed on glasses slightly varies between two glasses successively produced, flaw detection by color image analysis is a challenging problem. In order to overcome this problem, the aspect flaws detection is based on an original color image segmentation scheme that iteratively constructs pixel classes. Since the results of segmentation scheme are known to depend on the choice of the color space, the main originality of our approach is to automatically determine the most discriminant color space for each class to be constructed. Experimental results show that the selection of the well-suited color spaces contributes to improve the segmentation accuracy and the aspect flaw detection.

Object tracking is considered to be a key technique in many computer vision applications, such as video surveillance, object recognition, and robotics. We propose a method that improves the performance of scale-invariant feature transform (SIFT)-based object tracking algorithm to track the object in the subsequent video frames. Recently, many feature-based tracking methods have been proposed. An efficient and improved SIFT feature matching-based tracking method via neural network is provided and compares the outcome of this method with other tracking method outcomes. The tracked object is assigned a distance with the Kinect sensor to determine the depth of the detected object. The experimental results show that the proposed method can track the target object under different situations such as rotation, scaling, and many others with less computation time. Self-organizing map-based improved object tracking method can also estimate the distance between the tracked object and the image sensor. The proposed tracking technique will be useful for the development of many computer vision and robot navigation applications.

Fractional pixel motion estimation (ME) is required to achieve more accurate motion vectors and higher compression efficiency. This results in an increase in the computational complexity of the ME process because of additional computational overheads such as interpolation and fractional pixel search. Fast algorithms for fractional ME in H.264/AVC are presented. To reduce the complexity of fractional pixel ME, unimodal error surface assumption is used to check only some points in the fractional pixel search window. The proposed algorithm employs motion prediction, directional quadrant and point-based search pattern and early termination to speed up the process. Hence, the proposed algorithm efficiently explores the neighborhood of integer pixel based on high correlation that exists between the neighboring fractional pixels and unimodal property of error surface. The proposed search pattern and early termination reduce computational time by almost 8% to 18% as compared to the hierarchical fractional pixel algorithm employed in the reference software with a negligible degradation in video quality and negligible increase in bit rate.

An adaptive total variation method to reduce speckles with preservation of targets in synthetic aperture radar (SAR) images is investigated. Based on the gamma distribution of speckle, an adaptive total variational model is proposed with its fidelity term derived from a framework of weighted maximum likelihood estimation and its regularity term with constraints on the gradient of an image. It has merits of preserving textures and targets since the a priori distribution of noise is incorporated into the model and the weights are essentially image data driven, which can adaptively adjust the weights. The mathematical analysis is carried out, and proof of existence and uniqueness of a solution for the corresponding function is also presented. Theoretical analysis and experiments on both the simulated and real SAR images demonstrate that the method proposed here performs favorably.

Optimal texture analysis algorithms for describing degradation of carpets are identified. Experimental design is applied to select from a set of texture analysis algorithms those optimal for identifying texture changes due to degradation of carpets. The degree of wear of a degraded carpet is quantified by comparing its texture to the original texture. The set of texture algorithms is applied on intensity images obtained from the American and the European standards. The performance of the texture algorithms is evaluated using measures that quantify characteristics in the relationship between the metrics and the changes in texture. The statistical analysis of the experimental results shows that the local binary patterns algorithm is optimal in <50% of the cases, for describing degradation of the carpets. Other texture algorithms that optimally characterize the degradation of carpets include the use of the power spectrum, Wigner distribution, and average co-occurrence matrix algorithms.

In order to improve discriminative power and illumination invariance, local color descriptors have been proposed. However, the high dimensionality of color descriptor limits the performance of matching in terms of speed. It is desirable to reduce the dimensionality for efficient computation and compact representation. A compaction of color descriptor is presented. The color feature vector is represented as a quaternion vector and then projected into a precomputed eigenspace. The resulting descriptor is highly compact and the experimental results show that the performance is almost identical to that of the standard RGB-SIFT.

The classical circularity measure that considers the relationship between the shape area and the shape perimeter is sensitive to noise and limited for shapes with boundary defects (which lead to a large increase in perimeter). We propose a recent and generic method for shape circularity measure. The proposed method derives from radial moments and satisfies the desirable properties of circularity measures. Compared with the classical circularity measure and Hu moment invariants based method, the new measure is robust to noise and performs better in the case of shapes with boundary defects. Experimental results also show the superiority of the proposed method.

Active shape model (ASM) is a statistically parametrical model. It is widely used for facial feature extraction in face images. An ASM method is proposed. First, intensity values used in original ASM cannot provide enough information for model searching, which is sensitive to lighting conditions and so on. So we use a measure which indicates the orientation of structure at each pixel instead of intensity value to represent image texture. In addition, a new method is adopted for building a local profile model, which makes full use of texture information around the landmarks. Experimental results show that the improved ASM can locate face features more accurately than other ASM methods, and our method is more robust to poses, expression, and illumination variations.

Human age estimation in face images has become an active research topic in computer vision because of many potentially useful applications. However, the face may be unavailable or unusable in reality, especially for noncooperative users. So, a question is raised: Is it still possible to extract some relevant information on age when the face cannot be used? A new idea for age classification based on human body anthropometry is proposed which does not rely on the face. Based on this idea, a computational algorithm is developed to separate children from adults using relative lengths of body parts. To demonstrate the feasibility of this technique, the proposed age classification method is applied to two public databases of articulated human body images. A good accuracy is achieved on both databases, which is very encouraging.

For accurate color reproduction of motion pictures or still pictures of moving objects, we developed a different one-shot six-band image-capturing and visualization system that combines multiband and stereo imaging techniques. The proposed system consists of two consumer-model digital cameras and an interference filter whose spectral transmittance is comb-shaped. Subpixel correspondence search between the stereo image pairs is conducted and image transformation whose parameters are estimated from the correspondence is used to correct the geometric relationship between the images. The Wiener estimation method is used for color reproduction. For experiments, we have constructed two six-band camera systems to evaluate the quality of resultant images. One is for capturing high-resolution images using digital single-lens reflex cameras. The other is for capturing motion pictures using digital video cameras, for which all image processing steps after image capture are implemented on graphics processing units and the frame rate of the system is 30 fps when image size is XGA. For both systems, the average color difference between the measurement data and estimation results for 24 color patches of the Macbeth ColorChecker™ is dEa ∗ b ∗ =1.21 at maximum.