Search

In error-prone environments, packets could be lost during the transmission of compressed video bitstream, and corrupted errors by the lost packets could be propagated until it meets a refresh point. To defect the error propagation, constrained intra-prediction in high efficiency video coding (HEVC) performs intra-prediction using only neighboring intra-coded samples. However, the errors could be propagated to intra-coded blocks by an in-loop filtering process, since the in-loop filtering is performed regardless of the prediction modes of blocks, even though constrained intra-prediction is used. In this paper, a constrained in-loop filtering is proposed to protect intra-coded samples from error propagation by adaptively applying filters depending on the prediction modes of reconstructed blocks. Simulation results show that the proposed method alleviates the distortion of the decoded pictures and thereby improves both subjective and objective quality compared with the HEVC standard.

We propose a real-time pose estimation method that addresses the weaknesses of the numerical inverse kinematics method. Using conventional inverse kinematics based on the numerical method requires many iterations; moreover, a singularity in the Jacobian matrix as well as a local minimum problem can occur. To solve these problems, we propose an inverse kinematics method combined with an unscented Kalman filter (UKF) to recover intermediate joint information. Because the numerical inverse kinematics method optimizes a state, the solution can often converge to the local minimum and require many iterations. We use several sigma points for analysis to find the optimum state by using an unscented transform. The improved method using a UKF converges faster than the numerical inverse kinematics method for the global minimum of the existing inverse kinematics. We use 2-D image processes to extract body areas from the input images, and a 3-D reconstruction algorithm is used to estimate the 3-D positions of the extracted human body area. Using the improved method, we generate intermediate joints for each body part and the results show that the proposed method reduces the computational complexity and increases the accuracy of estimation compared to conventional numerical inverse kinematics.

A multiview video-based three-dimensional (3-D) video system offers a realistic impression and a free view navigation to the user. The efficient compression and intermediate view synthesis are key technologies since 3-D video systems deal multiple views. We propose an intermediate view synthesis using a rectangular multiview camera system that is suitable to realize 3-D video systems. The rectangular multiview camera system not only can offer free view navigation both horizontally and vertically but also can employ three reference views such as left, right, and bottom for intermediate view synthesis. The proposed view synthesis method first represents the each reference view to meshes and then finds the best disparity for each mesh element by using the stereo matching between reference views. Before stereo matching, we separate the virtual image to be synthesized into several regions to enhance the accuracy of disparities. The mesh is classified into foreground and background groups by disparity values and then affine transformed. By experiments, we confirm that the proposed method synthesizes a high-quality image and is suitable for 3-D video systems.

We propose a rate-distortion optimized transform coding method that adaptively employs either integer cosine transform that is an integer-approximated version of discrete cosine transform (DCT) or integer sine transform (IST) in a rate-distortion sense. The DCT that has been adopted in most video-coding standards is known as a suboptimal substitute for the Karhunen-Loeve transform. However, according to the correlation of a signal, an alternative transform can achieve higher coding efficiency. We introduce a discrete sine transform (DST) that achieves the high-energy compactness in a correlation coefficient range of −0.5 to 0.5 and is applied to the current design of H.264/AVC (advanced video coding). Moreover, to avoid the encoder and decoder mismatch and make the implementation simple, an IST that is an integer-approximated version of the DST is developed. The experimental results show that the proposed method achieves a Bjontegaard Delta-RATE gain up to 5.49% compared to Joint model 11.0.

Fundamental-matrix and key-frame selection constitute one of the most important techniques for full 3-D reconstruction of objects from turntable sequences. This paper proposes the new methods for these selection problems in 3-D reconstruction from uncalibrated sequences taken with a turntable and the Fotonovo camera system. Also, we propose a projection-matrix refinement for accurate full 3-D reconstruction. Our approach utilizes single-axis motion. To evaluate the fundamental matrix, camera calibration and 3-D registration are generally needed. Our main contribution is a method for robustly determining the corresponding points between two images, and for accurately filling gaps in a sparse object so as to make surface reconstruction tractable. We do not need all frames, but only few pairs of images (key frames). The key-frame selection has the advantage in camera pose estimation and 3-D scene reconstruction of reducing the computational costs. Experimental results on real image sequences demonstrate accurate object reconstructions and robustness of the proposed methods.

An essential role of model-based vision deals with the problem of solving for the values of all viewpoint and model parameters to be computed for the dimensions of an object for a three-dimensional (3-D) model. We propose a new method using only three vanishing points for recovering the dimensions of an object, its position, and the focal length of the camera from a single image of weak perspective projection. Our aim is to compute the dimension vector v of a model represented by the unit vector of objects from the image. The optimization of the dimension vector v for the object can be thus solved by the standard nonlinear optimization techniques with a multistart method that generates multiple starting points for the optimizer by sampling the parameter space uniformly. Experimental results demonstrate the dimension vector v of the proposed method for the 3-D model and show that the performance of the proposed method is better than the conventional. We also address that the actual dimensions of object from the image agree well with the simulated results from the single image of weak perspective projection.

Camera calibration is essential in order to perform 3-D reconstruction from uncalibrated images. This paper proposes a new technique of camera calibration using a homography among three planar patterns taken from a single image. It is implemented more easily and simply than conventional methods. Experimental results support the hypothesis and show that the performance of the proposed method is better than that of conventional methods. Examples of 3-D reconstruction are demonstrated using the proposed algorithm on sequences of images.

We propose a two-color infrared counter-countermeasure (IRCCM) in a crossed-array tracker for efficient target tracking under infrared countermeasure (IRCM) conditions. The proposed two-color IRCCM uses two detection bands in the mid and near IR regions for target and IRCM detections, respectively. The ratio of the incident signals passing the two detection bands is calculated, and a threshold ratio is determined to distinguish the target signal from the target-IRCM mixed signal, i.e., to generate a target pulse signal without IRCM effects. Our IRCCM also adopts target position prediction to make provision against wrong target detection. To evaluate the performance of the proposed IRCCM, we construct a simulator and perform target tracking under various conditions. The simulation results show that the proposed IRCCM allows efficient target tracking by minimizing the IRCM effects.

There have been many techniques that manipulate a compressed image in the discrete cosine transform (DCT) domain for MPEG and JPEG standards, and we have proposed a model-based discontinuity evaluation technique in the DCT domain that has problems with rotated or nonideal discontinuities. To enhance the effectiveness of the discontinuity detection method in the DCT domain while preserving the advantage of efficiency, we propose a fuzzy filtering technique that consists of height fuzzification, direction fuzzification, and fuzzy filtering of discontinuities. The enhancement achieved by the fuzzy filtering includes the linking, thinning, and smoothing of discontinuities in the DCT domain. Although the detected discontinuities are rough in a low-resolution image for the size (8×8 pixels) of the DCT block, experimental results show that this technique is fast and stable to enhance the quality of discontinuities.

We propose a reformed method that utilizes the motion vectors (MVs) in an MPEG sequence as the motion depicter for motion analysis and representation of video contents. The MVs are converted to a uniform MV set, independent of the frame type and the direction of prediction, and then used as motion depicters in each frame. To obtain such a uniform MV set, a new motion analysis method using bidirectional prediction-independent framework is proposed. Generally, it is impossible to directly compare an I frame without MV to others such as B or P frames. But, this approach enables a frame-type-independent representation that normalizes temporal features including frame type, macroblock (MB) encoding, and MVs. Experimental results show that our method has good performance and high validity. Compared with a full-decoding method, the average of the processing time in our method is reduced about 55%, because our method is directly processed on the MPEG bit stream after variable length code (VLC) decoding. Average of the effective number of the normalized MVs in the proposed algorithm is increased about 25% than that of the conventional method.

We propose a new algorithm that generates intermediate scenes from the near and the far views. Assuming that two cameras can be approximated with the affine model, we align the optical axes of the near and the far views on the transition baseline using the Fourier method. Then we determine the scale of the desired novel scene, zoom in the far view, and zoom out the near view. The zoom-in far view is theoretically the same as the zoom-out near view, but disparities exist that cause the synthesized image to present the double-blurring effect. Therefore, we remove the disparities of the zoom-out near view and then synthesize it with the zoom-in far view. For the panorama-based navigation system, the proposed algorithm can also generate images having smooth and realistic transitions as the viewer walks from the current panoramic image to the next panoramic image.

The characteristics of the use of a hollow-core splice scheme for macrobend measurements are discussed both theoretically and experimentally. A perturbation theory for the modes of a bent hollow-core fiber and its loss characterization are developed so as to better understand the characteristics of the scheme. The maximal detection range of fabricated sensors with the proposed scheme is experimentally determined to be as large as a few hundred millimeters relative to the radius of curvature. In addition, the numerical estimation of the loss characteristics using the scheme shows modal trends which are in good agreement with experimental data.

We propose an eminent scene change detection (SCD) algorithm, and also present a new video-indexing method for fast content- based browsing and retrieval in video databases. The first step in our approach is to extract representative frames by using the proposed SCD algorithm from the Motion Picture Experts Group (MPEG) video sequence. Thereafter, we perform the video indexing by applying the rosette pattern (RP) to the extracted representative frames, and retrieve them. A principal cause of these processing procedures is that the performance of video indexing and retrieval system depends on the SCD. And how accurately SCD performs as well as a preprocessing are some of the considerations that we must take into account. Therefore, we classify every possible scene change that may happen within the video into three types: abrupt, gradual, and flashlight. And we detect every kind of scene change. Our SCD algorithm is better than the conventional ones in terms of SCD performance and we demonstrate its superiority by experiments. We also extract representative frames that represent series of detected scenes by the proposed accumulative histogram intersection measure (AHIM). Finally, we apply the RP to extracted representative frames for indexing, we can remarkably reduce the number of pixels (NOP) required to index and excellently retrieve the video scene.

We propose a new two-color cancellation algorithm as an in- frared (IR) counter-countermeasure technique for efficient target tracking under IR countermeasures, such as flares. For the proposed algorithm, we use two different spectral bands in the middle and near IR regions and define the ratio between the two bands' signal (MNR) to estimate the spectral distribution characteristics of the target and the flare. The pro- posed algorithm computes the MNR of the band signals, detects the MNR of the flare (MNR F) from the MNR histogram, and then cancels the flare signal from the middle band signal using the MNRF. We simulate the MNRF detection and the target signal extraction for the performance evaluation. The simulation results confirm that the proposed algorithm can cancel efficiently the effect of IR countermeasures under various conditions. Particularly, unlike conventional algorithms, the proposed al- gorithm can extract very well the target signal even when the target and flare signals are mixed in time domain. In addition, it is applicable to any ever-changing environment. © 2001 Society of Photo-Optical Instrumentation En- gineers. (DOI: 10.1117/1.1386925)

Conventional stationary reticle seeker uses an analog signal processing technology to extract a target position in the field of view. The analog technology is optimized for only frequency modulation signals, which makes it difficult to use various and adaptive signal processing, and causes the performance limit of the conventional technology. We propose a novel adaptive digital signal processing algorithm to overcome the performance limit. The proposed algorithm classifies the modulation signals into the FM and FM/AM signals and applies a different signal processing algorithm, which is optimized for the modulation characteristics of each signal. The algorithm also employs an adaptive clipping level to minimize the noise effect. Our simulation results show that the proposed algorithm is superior to the conventional one in terms of the linearity of the static gain curve, tracking performance, and robustness against noise.

To extract a precise target position in the presence of infrared countermeasures and significant noise, we propose a new algorithm based on the instantaneous frequency (IF) estimation of the sinusoidal frequency modulation (SFM) signals for a fixed-reticle seeker. To esti- mate the IF for target position extraction, the proposed algorithm maps the multicomponent SFM signal into the time-frequency plane to exam- ine its time-frequency distribution, and then applies the Hough transform. The simulation results show that the proposed algorithm can extract the target position efficiently in the presence of flare countermeasures and noise. © 2000 Society of Photo-Optical Instrumentation Engineers. (S0091-3286(00)00209-9)

Hyun-Ki HongChung-Ang UniversityGraduate School of Advanced ImagingScience, Multimedia, and FilmDepartment of Image EngineeringImage and Information Laboratory221 Heuksuk-Dong, DongJak-KuSeoul 156-756, South KoreaE-mail: jsk@candy.ee.cau.ac.krDong-Sun SeoMyong-Ji UniversityDepartment of Electronic EngineeringYonginKyonggido 449-728, South KoreaJong-Soo Choi

Reticle systems, which are widely used in infrared (IR) missile seekers, are considered to be the classical approach for estimating the position of a target in the field of view (FOV). A new simulation tool is proposed that gives tracking results of the concentric annular ring reticle seeker in various cases. Our simulation model is applicable to the devel- opment of the advanced seekers. When false targets such as flares are present in the FOV, simulation results show that the existing seeker cannot determine the precise target location. To decrease the suscepti- bility to countermeasures such as flares, we propose an efficient counter-countermeasure using signal prediction and correlation. In the simulation results, we ascertain that with our technique the reticle seeker can perform more effective target tracking than previous seekers. © 1997 Society of Photo-Optical Instrumentation Engineers. (S0091-3286(97)03211-X) Subject terms: reticle seeker; simulation; countermeasure; counter- countermeasure; signal prediction; correlation.

A dynamic simulator of reticle seekers is developed to predict tracking performance in various engagement scenarios. We can analyze the performance of the infrared (IR) reticle seekers with targets in vari- ous positions in the field of view and with several SNRs. We show that the presented simulator is a very powerful means for developing signal processing algorithms in the presence of IR countermeasures (IRCM). A simple and efficient algorithm for IR counter-countermeasures (IRCCM) is proposed and simulated. With the incorporation of this IRCCM tech- nique into the reticle seeker in the presence of IRCM, the missile can intercept its target successfully. © 1997 Society of Photo-Optical Instrumentation Engineers. (S0091-3286(97)00908-2)

Reticle seekers temporally modulate target location onto the incoming spatial signal. When large or multiple targets are present in the field of view (FOV), however, it is hard to precisely modulate the incom- ing target signal by the reticle. To solve this loss-of-modulation-depth problem, we present an improved reticle seeker using the segmented focal plane array (FPA). The new reticle system uses the normalized difference as well as the modulated signal of each detector output in the segmented FPA. In simulation, we have ascertained the proposed sys- tem can perform effective analysis and tracking for multiple or large tar- gets. © 1997 Society of Photo-Optical Instrumentation Engineers. (S0091-3286(97)02803-1)

We investigate edge extraction and the segmentation of range images. We first discuss problems that arise in conventional region-based and edge-based segmentation methods, then we propose a new edge-based segmentation algorithm. In the proposed algorithm, we extract edge contours by using the Gaussian filter and directional derivatives and segment a range image by integrating detected roof and step edges. Thus the correspondence between step and roof edges is not necessary. Computer simulations with several range images show that the proposed edge-based range image segmentation method is simpler and yields better performance than the conventional methods.