A RST-Invariant Robust DWT-HMM Watermarking Algorithm Incorporating Zernike Moments and Template

The design and implementation of a single instruction multiple data depth-from-motion algorithm on the image understanding architecture simulator are described. Correspondences are established in parallel for two temporarily separated images through correlation. The correspondences are used to determine the translational and rotational motion parameters of the camera through a parallel motion algorithm. This is done by first determining the appropriate translational parameters and then constraining the search for the exact translational and rotational parameters. The dense depth map is computed from the image correspondences and the computed motion parameters. Results are analyzed for three image sequences acquired from mobile vehicles. Depths are obtained at an average accuracy of about 8% in outdoor image sequences. The depth maps are processed to locate relatively small obstacles, like cans and cones, to a distance of about 60 ft. Large obstacles, like hills, are located even when they are much further away. >

Robustness is the key issue in the development of multi-bit watermarking algorithm. A new algorithm for robust multi-bit image watermarking based on Hidden Markov Model (HMM) in wavelet domain is proposed in this paper. The algorithm is characterized as follows: (1) the proposed blind detector based on vector HMM, which describes the statistics of wavelet coefficients, achieves significant improvement in performance compared to the conventional correlation detector; (2) adaptive watermark embedding scheme is applied to achieve the low distortion according to the Human Visual System (HVS); (3) optimal multi-bit watermark embedding strategy and maximum-likelihood detection for tree structure of vector HMM is proposed through system robustness analysis. Simulation results show that relatively high capacity for watermark embedding in low frequency subbands of wavelet domain is achieved with the proposed algorithm, and high robust results are observed against StirMark attacks, such as JPEG compression, additive noise, median cut and filter.

We aim at segmenting small organs (e.g., the pancreas) from abdominal CT scans. As the target often occupies a relatively small region in the input image, deep neural networks can be easily confused by the complex and variable background. To alleviate this, researchers proposed a coarse-to-fine approach [46], which used prediction from the first (coarse) stage to indicate a smaller input region for the second (fine) stage. Despite its effectiveness, this algorithm dealt with two stages individually, which lacked optimizing a global energy function, and limited its ability to incorporate multi-stage visual cues. Missing contextual information led to unsatisfying convergence in iterations, and that the fine stage sometimes produced even lower segmentation accuracy than the coarse stage. This paper presents a Recurrent Saliency Transformation Network. The key innovation is a saliency transformation module, which repeatedly converts the segmentation probability map from the previous iteration as spatial weights and applies these weights to the current iteration. This brings us two-fold benefits. In training, it allows joint optimization over the deep networks dealing with different input scales. In testing, it propagates multi-stage visual information throughout iterations to improve segmentation accuracy. Experiments in the NIH pancreas segmentation dataset demonstrate the state-of-the-art accuracy, which outperforms the previous best by an average of over 2%. Much higher accuracies are also reported on several small organs in a larger dataset collected by ourselves. In addition, our approach enjoys better convergence properties, making it more efficient and reliable in practice.

The ideal of Bessel-Fourier moments (BFMs) for image analysis and only rotation invariant image cognition has been proposed recently. In this paper, we extend the previous work and propose a new method for rotation, scaling and translation (RST) invariant texture recognition using Bessel-Fourier moments. Compared with the others moments based methods, the radial polynomials of Bessel-Fourier moments have more zeros and these zeros are more evenly distributed. It makes Bessel-Fourier moments more suitable for invariant texture recognition as a generalization of orthogonal complex moments. In the experiment part, we got three testing sets of 16, 24 and 54 texture images by way of translating, rotating and scaling them separately. The correct classification percentages (CCPs) are compared with that of orthogonal Fourier-Mellin moments and Zernike moments based methods in both noise-free and noisy condition. Experimental results validate the conclusion of theoretical derivation: BFM performs better in recognition capability and noise robustness in terms of RST texture recognition under both noise-free and noisy condition when compared with orthogonal Fourier-Mellin moments and Zernike moments based methods.

This paper presents a simple and analytical procedure for calibrating extrinsic camera parameters. First, a calibration equation that separates rotational and translational parameters is given. The calibrating equation involves only rotational parameters and requires no absolute position information. A four-point calibration procedure is proposed that involves three points on a line and one point out of the line, and leads to four possible calibration solutions, obtained analytically. Additional steps required to eliminate false solutions are also discussed. Once the true rotational parameters are identified, the translational parameters are obtained analytically and uniquely with additional absolute position information. The required absolute position information appears in a simple and explicit form. For extrinsic calibration among multiple cameras, such as stereo cameras, it is easy to show that the absolution position information is not needed.

It is essential to align different 3d models from different scale, posture and translation to a uniform coordinate system in many 3d applications. Traditional alignment algorithm iterative optimizes the scale, posture and translation parameters from random initial values which is usually time consumed especially dealing with huge 3d point cloud data. To solve this problem, a novel alignment algorithm is proposed, which mainly consists of two step. At the first step, the scale, posture and translation are quickly adjust by re-projecting and least square solving. At the second step, the scale, posture and translation parameters are fine tuned by iterative optimization. The experiments show that the alignment algorithm is efficient and accurate.

A novel method to recognize occluded objects using Markov model is proposed in this paper. In addition to Markov model having a high tolerance to noise, spatial distribution of features can be incorporated into Markov model in a natural and elegant way. Thus, high recognition accuracy can be achieved by the proposed method. More specifically, for each occluded object in the scene image, its translation, rotation and scale parameters can all be determined by our method even when it may have transformation parameters different from others or it may be duplicated in the scene image with transformation parameters different from each other. Moreover, the recognition process can be performed step by step to find out all of the objects in the scene image according to the confidence measure. Finally, the recognition process can be terminated automatically without knowing the number of objects included in the scene image since hypothesis verification and termination test are performed in our method. Actually, the solution of our method is useful for depth-search applications such as inspection of printed circuit board with multiple layers, underwater diving for searching objects, underground drilling for exploring mine, etc. The proposed method has been applied on two types of databases: puzzle and tool. The effectiveness and practicability of the proposed approach have been proven by various experimental results.

The easiest and most commonly used method to register the MRI and MEG datasets is point matching, where a few ‘fiducial’ points, i.e. nasion, 2 preauriculars, are marked during the acquisition of either technique. Another methodology involves the use of a stereotactic frame. In either case, the registration of the two modalities requires the definition of the correct transformation matrix with respect to scaling, translation and rotation parameters. We have exploited both the above strategies and, in general, root mean square (RMS) mismatches per point of less than 1 mm (simulations) or 6 mm (general practice with real data) were achieved.

Traditionally, the coarse and fine mechanical stages within a dual-stage positioning system are designed separately without consideration about their control performance in dual-stage integration. In this paper, a flexure-based Lorentz motor fine stage is designed concurrently with a simple PID controller to perform the dual-stage positioning, based on the existing coarse stage. The design of fine stage is carried out using a proposed integrated servo-mechanical design approach, where various specifications are considered and formulated as constraints in an optimization problem. Through the approach, both the plant modal parameters and controller parameters of fine stage are concurrently solved. It demonstrates that through suitable mechanical plant design, the fine stage can fulfill various control specifications by only using a simple PID controller, e.g., to compensate sensitivity peak, maintain stability, etc. Meanwhile, the proposed design approach also ensures certain open-loop positioning performance and the decoupling property of coarse/fine stages. The prototype of designed fine stage is fabricated, and experimental investigation indicates that the sensitivity peak is effectively reduced from 14.5 dB of coarse stage to 7.1 dB in the dual-stage system, and the fine stage is able to achieve submicrometer accuracy. The maximal tracking error is also reduced significantly from about 20 μm via the coarse stage to less than 2 μm through dual-stage positioning.

Bone suppression of chest radiographs (CXRs) is potentially useful for diagnosing lung diseases for radiologists and computer–aided diagnosis. This paper presents a cascaded convolutional network model in wavelet domain (Wavelet-CCN) for bone suppression in single conventional CXR. Wavelet coefficients are sparse and suitable as the output of convolutional network. The convolutional networks are trained to predict the wavelet coefficients of bone images from the wavelet coefficients of CXRs, using real two-exposure dual energy subtraction (DES) CXRs as training data. By combining the multi-level wavelet decomposition and a cascaded refinement framework, the Wavelet-CCN model can work automatically with a multi-scale approach and progressively refine the prediction in terms of accuracy and spatial resolution. Compared with previous work of CamsNet model which preforms bone prediction in gradient domain, the Wavelet-CCN model predicts the wavelet coefficients to reconstruct bone images and can avoid the inconsistent background intensity caused by 2D integration of gradients. The predicted bone image is subtracted from the original CXR to produce a soft-tissue image. The Wavelet-CCN model and its variants with different wavelet basis are evaluated on a dataset that consists of 504 cases of real two-exposure DES CXRs (404 cases for training and 100 cases for test). Experimental results show that among all the variants and different wavelet bases, the Wavelet-CCN model with Haar wavelet performs best. The average peak signal-to-noise ratio and structural similarity index of the soft-tissue images produced by the proposed Wavelet-CCN model are both higher than those of the previous CamsNet model in gradient domain, reaching values of 39.4 (±0.94) dB and 0.977 (±0.004), respectively. The results also demonstrate that the Wavelet-CCN model can process the CXRs acquired by four types of X-ray machines.

In this study, we propose a nonparametric simultaneous test procedure for the location translation and scale parameters. We consider the Wilcoxon rank sum test for the location translation parameter and the Mood test for the scale parameter with the quadratic and maximal types of combining functions. Then we derive the limiting null distributions of the combining functions. We illustrate our procedure with an example and compare efficiency by obtaining the empirical powers through a simulation study. Finally, we discuss some interesting features related to the nonparametric simultaneous tests.

Point set registration is an essential technique in the field of machine vision. In this article, we propose a robust global optimal solution to for the point set registration of feature points extracted from visual images, used in remote (300–120 km) space target tracking and targeting tasks. Specifically, we begin with cases where correspondences among point sets are known, establishing a cost function centered on maximizing the consensus set, wherein rotational and translational parameters are determined using voting methods and the branch-and-bound (BnB) algorithm, respectively. We then adapt this foundation to tackle the more challenging scenario of unknown correspondences in simultaneous pose and correspondence registration by adjusting the cost function and BnB bounding functions, supplemented with nested iterations to accurately determine rotation and translation parameters. Finally, the comprehensive experimental comparisons executed across synthetic and real datasets, along with ground-based spacecraft pose measurement setup, illustrate that, compared to existing methods, our proposed approach achieves precise estimations under the influence of noise and outliers. Moreover, compared to the globally nested BnB scheme, our method reduces computational complexity and enhances solution speeds.

This paper describes an application of the EM (expectation and maximisation) algorithm to the registration of incomplete millimetric radar images. The data used in this study consists of a series of nonoverlapping radar sweeps. Our registration process aims to recover transformation parameters between the radar-data and a digital map. The tokens used in the matching process are fragmented line-segments extracted from the radar images which predominantly correspond to hedge-rows in the cartographic data. The EM technique models data uncertainty using Gaussian mixtures defined over the positions and orientations of the lines. The resulting weighted least-squares parameter estimation problem is solved using the Levenberg-Marquardt method. A sensitivity analysis reveals that the date-likelihood function is unimodal in the translation and scale parameters. In-fact the algorithm is only sensitive to the choice of initial rotation parameter; this is attributable to local suboptima in the log-likelihood function associated with /spl pi//3 orientation ambiguities in the map. The method is also demonstrated to be relatively insensitive to random measurement errors on the line-segments.

Precise segmentation of key tissues in medical images is of great significance. Although deep neural networks have achieved promising results in many medical image segmentation tasks, it is still a challenge for volumetric medical image segmentation due to the limited computing resources and annotated datasets. In this paper, we propose a multi-resolution coarse-to-fine segmentation framework to perform accurate segmentation. The proposed framework contains a coarse stage and a fine stage. The coarse stage with low-resolution data provide high semantic cues for the fine stage. Moreover, we embed active learning processes into coarse-to-fine framework for sparse annotation, the proposed multiple query criteria active learning methods can select high-value slices to label. We evaluated the effectiveness of proposed framework on two public brain MRI datasets. Our coarse-to-fine networks outperform other competitive methods under the condition of fully supervised training. In addition, the proposed active learning method only need 30% to 40% slices of one scan to produce relatively better dense prediction results than non-active learning method and one query criteria active learning methods.

High-precision stages require high-speed and high-precision control to improve their production throughput and quality. However, it is expected that their motion speed and accuracy will reach a limit in the near future if the structure of the conventional high-precision stage is used. Therefore, the authors designed and fabricated a stage called the catapult stage which has a decouplable structure consisting of a fine stage and a coarse stage. This stage is different from conventional dual stages in which the fine stage would be disturbed by the coarse stage since they contact with each other. This paper proposes a novel control system design for the catapult stage, and a settling time shortening control method using final-state control (FSC). So far, FSC is mainly applied to the applications such as hard disk drives whose initial states are the zero. However, it is important to consider the initial states for the catapult stage since the initial position, velocity and acceleration of the catapult stage are not equal to zero. Simulations and experimental results demonstrate the effectiveness of the proposed methods.