Search

We investigated whether genetic susceptibility to tuberculosis (TB) influences lung adenocarcinoma development among never-smokers using TB genome-wide association study (GWAS) results within the Female Lung Cancer Consortium in Asia. Pathway analysis with the adaptive rank truncated product method was used to assess the association between a TB-related gene-set and lung adenocarcinoma using GWAS data from 5512 lung adenocarcinoma cases and 6277 controls. The gene-set consisted of 31 genes containing known/suggestive associations with genetic variants from previous TB-GWAS. Subsequently, we followed-up with Mendelian Randomization to evaluate the association between TB and lung adenocarcinoma using three genome-wide significant variants from previous TB-GWAS in East Asians. The TB-related gene-set was associated with lung adenocarcinoma (p = 0.016). Additionally, the Mendelian Randomization showed an association between TB and lung adenocarcinoma (OR = 1.31, 95% CI: 1.03, 1.66, p = 0.027). Our findings support TB as a causal risk factor for lung cancer development among never-smoking Asian women.

Arsenic (As) exposure has been extensively studied by investigating As species (e.g., inorganic arsenic (iAs), monomethylarsonic acid (MMA), and dimethylarsinic acid (DMA)) in urine, yet recent research suggests that blood could be a possible biomarker of As exposure. These investigations, however, were conducted on iAs-contaminated areas, and evidence on populations exposed to low levels of iAs is limited. This study aimed to describe the levels and distributions of As species in urine and blood, as well as to estimate methylation efficiency and related factors in the Korean population. Biological samples were obtained by the Korean Ministry of Food and Drug Safety. A total of 2025 urine samples and 598 blood samples were utilized in this study. Six As species were measured using ultra-high-performance liquid chromatography with inductively coupled plasma mass spectrometry (UPLC-ICP-MS): As(V), As(III), MMA, DMA, arsenobetaine (AsB), and arsenocholine (AsC). Multiple linear regression models were used to examine the relationship between As species (concentrations and proportions) and covariates. AsB was the most prevalent species in urine and blood. The relative composition of iAs, MMA, DMA, and AsC in urine and blood differed significantly. Consumption of blue-backed fish was linked to higher levels of AsB in urine and blood. Type of drinking water and multigrain rice consumption were associated with increased iAs concentration in urine. Except for iAs, every species had correlations in urine and blood in both univariate and multivariate analyses. Adolescents and smokers presented a lower methylation efficiency (higher %MMA and lower %DMA in urine) and females presented a higher methylation efficiency (lower %iAs, %MMA, and higher %DMA in urine). In conclusion, blood iAs concentration cannot represent urinary iAs; nonetheless, different compositions of urine and blood might reflect distinct information about iAs exposure. Further investigations on exposure factors and health are needed using low-exposure groups.

One of the major challenges of model-free visual tracking problem has been the difficulty originating from the unpredictable and drastic changes in the appearance of objects we target to track. Existing methods tackle this problem by updating the appearance model on-line in order to adapt to the changes in the appearance. Despite the success of these methods however, inaccurate and erroneous updates of the appearance model result in a tracker drift. In this paper, we introduce a novel real-time visual tracking algorithm based on a template selection strategy constructed by deep reinforcement learning methods. The tracking algorithm utilizes this strategy to choose the appropriate template for tracking a given frame. The template selection strategy is self-learned by utilizing a simple policy gradient method on numerous training episodes randomly generated from a tracking benchmark dataset. Our proposed reinforcement learning framework is generally applicable to other confidence map based tracking algorithms. The experiment shows that our tracking algorithm runs in real-time speed of 43 fps and the proposed policy network effectively decides the appropriate template for successful visual tracking.

Increasing the depth of Convolutional Neural Networks (CNNs) has been recognized to provide better generalization performance. However, in the case of 3D CNNs, stacking layers increases the number of learnable parameters linearly, making it more prone to learn redundant features. In this paper, we propose a novel 3D CNN structure that learns shared 2D triplanar features viewed from the three orthogonal planes, which we term S3PNet. Due to the reduced dimension of the convolutions, the proposed S3PNet is able to learn 3D representations with substantially fewer learnable parameters. Experimental evaluations show that the combination of 2D representations on the different orthogonal views learned through the S3PNet is sufficient and effective for 3D representation, with the results outperforming current methods based on fully 3D CNNs. We support this with extensive evaluations on widely used 3D data sources in computer vision: CAD models, LiDAR point clouds, RGB-D images, and 3D Computed Tomography scans. Experiments further demonstrate that S3PNet has better generalization capability for smaller training sets, and learns more of kernels with less redundancy compared to kernels learned from 3D CNNs.

Novel large margin formulation for semantic similarity learning.Efficient optimization algorithm to solve the proposed semi-definite program (SDP).Thorough experimental study to compare the performances of several algorithms for hierarchical image classification.State-of-the-art classification performance under the hierarchical-loss criterion. In the present paper, a novel image classification method that uses the hierarchical structure of categories to produce more semantic prediction is presented. This implies that our algorithm may not yield a correct prediction, but the result is likely to be semantically close to the right category. Therefore, the proposed method is able to provide a more informative classification result. The main idea of our method is twofold. First, it uses semantic representation, instead of low-level image features, enabling the construction of high-level constraints that exploit the relationship among semantic concepts in the category hierarchy. Second, from such constraints, an optimization problem is formulated to learn a semantic similarity function in a large-margin framework. This similarity function is then used to classify test images. Experimental results demonstrate that our method provides effective classification results for various real-image datasets.

Under the popular Markov random field (MRF) model, low-level vision problems are usually formulated by prior and likelihood models. In recent years, the priors have been formulated from high-order cliques and have demonstrated their robustness in many problems. However, the likelihoods have remained zeroth-order clique potentials. This zeroth-order clique assumption causes inaccurate solution and gives rise to undesirable fattening effect especially when window-based matching costs are employed. In this paper, we investigate high-order likelihood modeling for the stereo matching problem which advocates the dissimilarity measure between the whole reference image and the warped non-reference image. If the dissimilarity measure is evaluated between filtered stereo images, the matching cost can be modeled as high-order clique potentials. When linear filters and nonparametric census filter are used, it is shown that the high-order clique potentials can be reduced to pairwise energy functions. Consequently, a global optimization is possible by employing efficient graph cuts algorithm. Experimental results show that the proposed high-order likelihood models produce significantly better results than the conventional zeroth-order models qualitatively as well as quantitatively.

We propose a novel deep learning based system for vessel segmentation. Existing methods using CNNs have mostly relied on local appearances learned on the regular image grid, without consideration of the graphical structure of vessel shape. Effective use of the strong relationship that exists between vessel neighborhoods can help improve the vessel segmentation accuracy. To this end, we incorporate a graph neural network into a unified CNN architecture to jointly exploit both local appearances and global vessel structures. We extensively perform comparative evaluations on four retinal image datasets and a coronary artery X-ray angiography dataset, showing that the proposed method outperforms or is on par with current state-of-the-art methods in terms of the average precision and the area under the receiver operating characteristic curve. Statistical significance on the performance difference between the proposed method and each comparable method is suggested by conducting a paired t-test. In addition, ablation studies support the particular choices of algorithmic detail and hyperparameter values of the proposed method. The proposed architecture is widely applicable since it can be applied to expand any type of CNN-based vessel segmentation method to enhance the performance.

In this paper, we present a new framework for three-dimensional (3D) reconstruction of multiple rigid objects from dynamic scenes. Conventional 3D reconstruction from multiple views is applicable to static scenes, in which the configuration of objects is fixed while the images are taken. In our framework, we aim to reconstruct the 3D models of multiple objects in a more general setting where the configuration of the objects varies among views. We solve this problem by object-centered decomposition of the dynamic scenes using unsupervised co-recognition approach. Unlike conventional motion segmentation algorithms that require small motion assumption between consecutive views, co-recognition method provides reliable accurate correspondences of a same object among unordered and wide-baseline views. In order to segment each object region, we benefit from the 3D sparse points obtained from the structure-from-motion. These points are reliable and serve as automatic seed points for a seeded-segmentation algorithm. Experiments on various real challenging image sequences demonstrate the effectiveness of our approach, especially in the presence of abrupt independent motions of objects.

Conventional particle filtering-based visual ego-motion estimation or visual odometry often suffers from large local linearization errors in the case of abrupt camera motion. The main contribution of this paper is to present a novel particle filtering-based visual ego-motion estimation algorithm that is especially robust to the abrupt camera motion. The robustness to the abrupt camera motion is achieved by multi-layered importance sampling via particle swarm optimization (PSO), which iteratively moves particles to higher likelihood region without local linearization of the measurement equation. Furthermore, we make the proposed visual ego-motion estimation algorithm in real-time by reformulating the conventional vector space PSO algorithm in consideration of the geometry of the special Euclidean group SE(3), which is a Lie group representing the space of 3-D camera poses. The performance of our proposed algorithm is experimentally evaluated and compared with the local linearization and unscented particle filter-based visual ego-motion estimation algorithms on both simulated and real data sets.

The pixel labeling problems in computer vision are often formulated as energy minimization tasks. Algorithms such as graph cuts and belief propagation are prominent; however, they are only applicable for specific energy forms. For general optimization, Markov Chain Monte Carlo (MCMC) based simulated annealing can estimate the minima states very slowly. This paper presents a sampling paradigm for faster optimization. First, in contrast to previous MCMCs, the role of detailed balance constraint is eliminated. The reversible Markov chain jumps are essential for sampling an arbitrary posterior distribution, but they are not essential for optimization tasks. This allows a computationally simple window cluster sample. Second, the proposal states are generated from combined sets of local minima which achieve a substantial increase in speed compared to uniformly labeled cluster proposals. Third, under the coarse-to-fine strategy, the maximum window size variable is incorporated along with the temperature variable during simulated annealing. The proposed window annealing is experimentally shown to be many times faster and capable of finding lower energy compared to the previous Gibbs and Swendsen-Wang cut (SW-cut) sampler. In addition, the proposed method is compared with other deterministic algorithms like graph cut, belief propagation, and spectral method in their own specific energy forms. Window annealing displays competitive performance in all domains.

Using the urinary biomarkers 1-hydroxypyrene-glucuronide (1-OHPG), 2-naphthol, and malondialdehyde (MDA), we evaluated seasonal and regional variations in polycyclic aromatic hydrocarbon (PAH) exposure and oxidative stress among Korean adults and children. In total, 322 children (175 male and 147 female) and 332 adults (47 male and 285 female) were recruited in two regions of Korea, one representing a metropolitan area (Seoul/Incheon) and the other an industrial (Pohang) area, from winter 2002 to spring 2003. The subjects voluntarily gathered their first morning urine void, which was immediately transported to our laboratory and stored at -20 °C. Urinary 1-OHPG was measured by synchronous fluorescence spectroscopy, 2-naphthol by HPLC, and urinary MDA by HPLC with a UV detector. The median urinary 1-OHPG concentration tended to be higher in the industrial region than in the metropolitan region (0.92 vs. 0.77 ng/mL; p=0.03), and higher in winter than in spring (0.95 vs. 0.73 ng/mL; p

It is becoming increasingly clear that there is wide heterogeneity in genetic predisposition to breast cancer and that breast cancer risk is determined by interactive effect between genetic and environmental factors.We investigated the combined effects of antioxidant vitamin intake and NOS3 genetic polymorphisms on breast cancer risk in a Korean population (Seoul Breast Cancer Study). Histologically confirmed breast cancer cases (n = 512) and age, menopause status-matched controls (n = 512) with no present or previous history of cancer were recruited from several teaching hospitals in Seoul during 2001-2003. Two genetic polymorphisms of NOS3 (298G T and -786 T C) were assessed by single base extension assays.No overall association between the individual NOS3 genotypes or diplotypes and breast cancer risk was found, although the difference between cases and controls in the frequency of the NOS3 894 G T polymorphism showed borderline significance (OR = 0.74, 95% CI = 0.52-1.06). There was no significant difference in energy intake or the intake of antioxidant vitamins between cases and controls, with the exception of vitamin E (OR = 0.49 lowest vs. highest quartile, P(trend) 0.01). On the other hand, our results suggest that antioxidant vitamin intake may modify the effects of the NOS3 -786 T C or 894 G T genetic polymorphisms on breast cancer risk. Although a multiplicative interaction was not observed, the protective effect of β-carotene intake on breast cancer risk was observed predominantly in individuals with the TG:TG diplotype of NOS3 (OR = 0.68) but not observed with others diplotype. An inverse association between vitamin E intake and breast cancer risk was observed for individuals with the NOS3 786 TC + TT genotype and the NOS3 894 GG genotype. In addition, folic acid had a protective effect in the NOS3 786 TT and NOS3 894 GT + TT genotype.Our results suggest that intake of antioxidant vitamins might modify the association between genetic polymorphisms of NOS3 and breast cancer risk.

In this paper, we present a new surfel (surface element) based multi-view stereo algorithm that runs entirely on GPU. We utilize the flexibility of surfel-based 3D shape representation and global optimization by graph cuts in the same framework. Unlike previous works, the algorithm is optimized to massive parallel processing on GPU. First, we construct surfel candidates by local stereo matching and voting. After refining the position and orientation of the surfel candidates, we extract the optimal surfels by employing graph cuts under photo-consistency and surfel orientation constraints. In contrast to the conventional voxel based methods, the proposed algorithm utilizes more accurate photo-consistency and reconstructs the 3D shape up to sub-voxel accuracy. The orientation of the constructed surfel candidates imposes an effective constraint that reduces the effect of the minimal surface bias. The entire processing pipeline is implemented on the latest GPU to significantly speed up the processing. The experimental results show that the proposed approach reconstructs the 3D shape of an object accurately and efficiently, which runs more than 100 times faster than on CPU.

We conducted a genome-wide association study of childhood acute lymphoblastic leukemia (ALL) in a case–control study conducted in Korea. Incident childhood ALL cases (n = 50) and non-cancer controls (n = 50) frequency-matched to cases by age and sex, recruited from three teaching hospitals in Seoul between 2003 and 2008, were genotyped using Affymetrix SNP Array 6.0 platform. ALL risks were estimated as odds ratios (ORs) and 95% confidence intervals (CIs) adjusted for age and birth weight. The false discovery rate (FDR) was used for adjusting multiple tests. Of these 1 million SNPs, six SNPs in 4 genes (HAO1 rs6140264, EPB41L2 rs9388856, rs9388857, rs1360756, C2orf3 12105972, MAN2A1 rs3776932) were strongly associated with childhood ALL risk (Pdominant ≤ 0.0001 and Ptrend < 0.006). These SNPs remained significant after FDR adjustment (FDR value

Genetic variation may be an important risk factor for multiple myeloma. A hallmark of tumor formation and growth is cell cycle dysregulation and apoptosis avoidance. We previously reported the association of genetic variation in caspase genes, the apoptotic-regulating family, and multiple myeloma risk. To further examine if genetic variation in key cell cycle and apoptosis genes alters multiple myeloma risk, we genotyped 276 tag SNPs in 27 gene regions in a population-based case–control study of non-Hispanic Caucasian women (108 cases; 482 controls) in Connecticut. Logistic regression assessed the effect of each SNP on multiple myeloma risk and the minP test assessed the association at the gene region level. Three gene regions were significantly associated with risk of multiple myeloma ( BAX minP = 0.018, CASP9 minP = 0.025, and RIPK1 minP = 0.037). Further explorations identified the most significant variant of BAX , RIPK1 , and CASP9 to be rs1042265, rs9391981, and rs751643, respectively. The A variant at rs1042265 (OR GA+AA = 0.40, 95% CI = 0.21–0.78) and the C variant at rs9391981 (OR GC+CC = 0.32, 95% CI = 0.12–0.81) were associated with a decreased risk of multiple myeloma. The G variant at rs7516435 was associated with an increased risk of multiple myeloma (OR AG = 1.48, 95% CI = 0.94–2.32; OR GG = 2.59, 95% CI = 1.30–5.15; p trend = 0.005). Haplotype analyses supported the SNP findings. These findings suggest that genetic variation in cell cycle and apoptosis genes may play a key role in multiple myeloma and warrant further investigation through replication studies.

In this paper, we propose a novel occlusion invariant face recognition algorithm based on Selective Local Non-negative Matrix Factorization (S-LNMF) technique. The proposed algorithm is composed of two phases; the occlusion detection phase and the selective LNMF-based recognition phase. We use a local approach to effectively detect partial occlusions in an input face image. A face image is first divided into a finite number of disjointed local patches, and then each patch is represented by PCA (Principal Component Analysis), obtained by corresponding occlusion-free patches of training images. And the 1-NN threshold classifier is used for occlusion detection for each patch in the corresponding PCA space. In the recognition phase, by employing the LNMF-based face representation, we exclusively use the LNMF bases of occlusion-free image patches for face recognition. Euclidean nearest neighbor rule is applied for the matching. We have performed experiments on AR face database that includes many occluded face images by sunglasses and scarves. The experimental results demonstrate that the proposed local patch-based occlusion detection technique works well and the S-LNMF method shows superior performance to other conventional approaches.

In this paper, we propose a new stereo matching algorithm using an iterated graph cuts and mean shift filtering technique. Our algorithm estimates the disparity map progressively through the following two steps. In the first step, with a previously estimated RDM (reliable disparity map) that consists of sparse ground control points, an updated dense disparity map is constructed through a RDM constrained energy minimization framework that can cope with occlusion. The graph cuts technique is employed for the solution of the proposed energy model. In the second step, more accurate and denser RDM is estimated through the disparity crosschecking technique and the mean shift filtering in the CSD (color-spatial-disparity) space. The proposed algorithm expands the reliable disparities in RDM repeatedly through the above two steps until it converges. Experimental results on the standard data set demonstrate that the proposed algorithm achieves comparable performance to the state-of-the-arts, and gives excellent results especially in the areas such as the disparity discontinuous boundaries and occluded regions, where the conventional methods usually suffer.

Objective . The aim of this study was to determine whether GSTM1 or GSTT1 might be associated with risk of cervical cancer development in Korean women. The multiplicative interaction of GSTM1 and GSTT1 genotype with p21 , p53 polymorphism, and HPV genotype was also investigated. Methods . From 1997 to 1999, uterine cervical carcinoma was diagnosed in 215 Korean women at the Department of Obstetrics and Gynecology of Seoul National University Hospital. None of the women in the control groups ( n = 98) had any evidence of cervical lesions, which were confirmed by Pap smear. Finally, 81 cases and 86 controls were genotyped for p21 , p53 , and GSTM1 and T1 and HPV infection. A multiplex PCR method was used for the genotyping of GSTM1 and GSTT1 ; direct sequencing for p53 codon 72, high-risk HPV, and PCR-RFLP (BsmAI) for p21 codon. The unconditional logistic regression analysis was used to calculate ORs and 95% CI. Results . Although the GSTM1 and GSTT1 genotype was not significantly associated with cervical cancer development for all women, the GSTM1 null genotype was significantly associated with an increased risk of cervical cancer development in women with high-risk HPV infection (OR = 2.9, 95% CI: 1.0–8.2). Although the frequency of overall GSTT1 null genotype was significantly lower in cervical carcinoma patients with high-risk HPV infection (OR = 0.3, 95% CI: 0.1–1.0), almost 2-fold increased risk was observed among women with GSTT1 null and Arg/Arg genotype (OR = 1.9, 95% CI: 0.7–5.4). Although the cervical cancer risk was 3.3-fold increased in women with null and Arg/Arg genotype compared to women with GSTM1 present and p21 Ser-containing genotype, there was no significant multiplicative interaction between GSTM1 and p21 ( P for interaction=0.785) or p53 ( P for interaction=0.815). Conclusion . These findings suggest that the risk of cervical cancer may be related to GSTM1 genotype in women with high-risk HPV infection and that there is a possible gene–gene interaction in the incidence of cervical cancer.

In this paper, we propose a progressive encoding algorithm for the geometric information of a 3D object, which is represented by binary voxels. Using the morphological pyramidal decomposition, the proposed algorithm first generates the multi-resolution models of a 3D object. Then, each resolution model is predicted from its lower resolution model, and the prediction errors are encoded using an arithmetic coding technique. To yield high compression ratio, each model is partitioned into the inside, boundary, and outside regions based on the lower resolution model. This partitioning method greatly reduces the amount of data to be encoded, since the prediction errors are compactly concentrated near the boundary region. Moreover, the neighborhood relation of each boundary voxel is used as the context for the arithmetic coding to further increase the compression efficiency. It is demonstrated by extensive simulation results that the proposed algorithm provides better coding gain than the conventional voxel and mesh compression algorithms.

In this paper, we propose a novel model-based perceptual grouping algorithm for the line features of 3-D polyhedral objects. Given a 3-D polyhedral model, perceptual grouping is performed to extract a set of 3-D line segments which are geometrically consistent with the 3-D model. Unlike the conventional approaches, grouping is done in 3-D space in a model-based framework. In our unique approach, a decision tree classifier is employed for encoding and retrieving the geometric information of the 3-D model. A Gestalt graph is constructed by classifying input instances into proper Gestalt relations using the decision tree. The Gestalt graph is then decomposed into a few subgraphs, yielding appropriate groups of features. As an application, we suggest a 3-D object recognition system which can be accomplished by selecting a best-matched group. In order to evaluate the performance of the proposed algorithm, experiments are carried out on both synthetic and real scenes.

To evaluate the potential association between NAT1/NAT2 polymorphisms and breast cancer, a case-control study was conducted in Korean women (254 cases, 301 controls). NAT1 *4/*10 genotype (42%) was the most common NAT1 genotype in this Korean population. The frequencies of slow, intermediate and rapid NAT2 acetylator genotype were 16, 39 and 44% in cases and 16, 42 and 42% in controls. Neither NAT1 rapid (homozygous or heterozygous NAT1 *10) (OR=1.2, 95% CI=0.8-1.9) nor NAT2 rapid acetylator genotype (OR=1.2, 95% CI=0.8-1.7) showed significant association with breast cancer risk. Although the risk of NAT2 rapid acetylator genotype in postmenopausal women (OR=1.4, 95% CI=0.7-2.8) was higher than that in premenopausal women (OR=1.1, 95% CI=0.7-1.7), those were not statistically significant. However, combinations of NAT1, GSTM1 and GSTT1 genotypes showed a significant linear gene-dosage relationship with breast cancer (p for trend=0.04) and those women with NAT2 rapid acetylator and both GSTM1 and GSTT1 null genotypes were at the elevated risk (OR=3.1, 95% CI=1.0-9.1). These results suggest that genetic polymorphisms of NAT1 and NAT2 have no independent effect on breast cancer risk, but they modulate breast cancer risk in the presence of GSTM1 and GSTT1 null genotypes.

In this paper, we propose a novel 2-D partial matching algorithm using the model-based probabilistic analysis of feature correspondences in the relation vector space, which is quite robust to shape variations due to noise and occlusions and invariant to 2-D geometric transformations as well. We represent an object using the attributed relational graph (ARG) model with nodes (features) of a set of the binary relation vectors. By defining relation vector space which can describe the structural information of an object centered at a specific feature, and modeling distortions due to partial occlusion or the input noise statistically in this space, lost features can be easily identified, so that the partial matching is performed efficiently. The proposed partial matching algorithm consists of two-phases. First, a finite number of candidate subgraphs are selected in an image, by using the logical constraint embedding local and structural consistency as well as the correspondence measure between model and image features. Second, the feature loss detection is done iteratively by the error detection and voting scheme through the error analysis in the relation vector space. Experimental results on real images demonstrate that the proposed algorithm has superior performance to those of the conventional relaxation algorithms, by localizing target objects robustly and correctly even in severely noisy and occluded scenes.

In this paper, we propose a new 3D target recognition algorithm using a single image. Our approach is based on geometrically invariant relationships. By employing a practical CCD camera model for projective image formation, and analyzing the constraints on the projection parameters, we derive two invariant relationships using six pairs of 3D space features and corresponding image features, such as points and lines. Compared to the conventional approach in which only a single invariant relationship is derived from six point pairs based on the general finite projective camera model, the two relationships obtained from the practical camera model can increase the effectiveness of matching. Based on the derived invariant relationships, a new view-invariant object recognition algorithm using a single image is proposed. The performance of the proposed recognition algorithm is demonstrated by various computer simulations on synthetic and real images of objects. The experimental results show that 3D objects in the image can be robustly recognized, using the linear features, by the proposed algorithm.

This paper presents new adaptive H.263+ rate control algorithms for video streaming and interactive video applications under networks supporting bandwidth renegotiation, which can communicate with end-users to accommodate their time-varying bandwidth requests during the data transmission. That is, the requests of end-users can be supported adaptively according to the availability of the network resources, and thus the overall network utilization can be improved simultaneously. They are especially suitable for the transmission of non-stationary video traffics. The proposed rate control algorithms communicate with the network to renegotiate the required bandwidth for the underlying video which are measured based on the motion change information, and choose their control strategies according to the renegotiation results. Unlike most conventional algorithms that control only the spatial quality by adjusting quantization parameters, the proposed algorithms treat both the spatial and temporal qualities at the same time to enhance human visual perceptual quality. Experimental results are provided to demonstrate that the proposed rate control algorithms can achieve superior performance to the conventional ones with low computational complexity under networks supporting bandwidth renegotiation.

In this paper, we derive new geometric invariants for structured 3D points and lines from single image under projective transform, and we propose a novel model-based 3D object recognition algorithm using them. Based on the matrix representation of the transformation between space features (points and lines) and the corresponding projected image features, new geometric invariants are derived via the determinant ratio technique. First, an invariant for six points on two adjacent planes is derived, which is shown to be equivalent to Zhu's result [1], but in simpler formulation. Then, two new geometric invariants for structured lines are investigated: one for five lines on two adjacent planes and the other for six lines on four planes. By using the derived invariants, a novel 3D object recognition algorithm is developed, in which a hashing technique with thresholds and multiple invariants for a model are employed to overcome the over-invariant and false alarm problems. Simulation results on real images show that the derived invariants remain stable even in a noisy environment, and the proposed 3D object recognition algorithm is quite robust and accurate.

The aim of motion stereo is to extract the 3-D information of an object from images of a moving camera using the geometric relationships between corresponding points. This paper presents an accurate and robust motion stereo algorithm employing multiple images, taken under a general motion. The object functions for individual stereo pairs are represented, with respect to the distance, then these object functions are integrated considering the position of cameras and the shape of the object functions. By integrating the general motion stereo images, we not only reduce the ambiguities in correspondence, but also improve the precision of the reconstruction. Also by introducing an adaptive window technique, we can alleviate the effect of projective distortion in matching features and improve the accuracy greatly. Experimental results on a synthetic and real data set are presented to demonstrate the performance of the proposed algorithm.

In this paper, we propose a new eigenvector-based line feature matching algorithm, which is invariant to the in-plane rotation, translation, and scale. First, in order to reduce the number of possible matches, we use a preliminary correspondence test that generates a set of finite candidate models, by restricting combinations of line features in the input image. This approach resolves an inherent problem relating to ordering and correspondence in an eigenvector/modal approach. Second, we employ the modal analysis, in which the Gaussian weighted proximity matrices for reference and candidate models are constructed to record the relative distance and angle information between line features for each model. Then, the modes of the proximity matrices of the two models are compared to yield the dissimilarity measure, which describes the quantitative degree of the difference between the two models. Experimental results for synthetic and real images show that the proposed algorithm performs matching of the line features with affine variation fast and efficiently and provides the degree of dissimilarity in a quantitative way.

In this paper, a new recognition algorithm for 2D object contours, based on the decimated wavelet transform, is presented, emphasizing the starting point dependency problem. The proposed matching algorithm consists of two parts: Firstly, we present new data structures for the decimated wavelet representation and a searching algorithm to estimate the misalignment between the starting points for the reference model and unknown object. We also adopt a polynomial approximation technique and propose a fast searching algorithm. And then, matching is performed in an aligned condition on the multiresolutional wavelet representation. By employing a variable-rate decimation scheme, we can achieve fast and accurate recognition results, even in the presence of heavy noise. We provide an analysis on the computational complexity, showing that our approach requires only less than 25% of the computational load required for the conventional method 1]. Various experimental results on both synthetic and real imagery are presented to demonstrate the performance of the proposed algorithm. The simulation results show that the proposed algorithm successfully estimates the misalignment and classifies 2D object contours, even for the input SNR = 5 dB.

Most conventional SFS (shape from shading) algorithms have been developed under three basic assumptions on surface properties and imaging geometry to simplify the problem, namely, a Lambertian surface, a distant point light source and orthographic projection. In this research, we derive a physics-based generalized reflectance map model which includes diffuse and specular reflection effects, a nearby point light source and perspective projection, and then we develop a new direct shape recovery algorithm from shaded images. The basic idea of our solution method is to discretize the image irradiance equation with a finite triangular element surface model, to express the resulting nonlinear system of equations in terms of depth variables only and to recover the object shape by linearizing the nonlinear equations and minimizing a quadratic cost functional. We perform numerical experiments with one or multiple photometric stereo images to demonstrate the performance of the derived physics-based reflectance map model and the proposed SFS algorithm.

Based on the traditional problem formulation, it is difficult to integrate the two important vision cues, i.e., shading and stereo, for shape reconstruction due to conflicting albedo and image projection assumptions. In this research, we propose a novel scheme to integrate shading and stereo. First, by using the photometric ratio, we derive a new SFS (shape from shading) formulation where no albedo is needed for shape reconstruction. Then, we establish a unified framework for the integration of photometric ratio and stereo by employing perspective projection on a parametric surface via minimizing a cost functional which consists of a weighted sum of shading and stereo errors. Simulation results are given to show the performance of our new robust algorithm.

Most conventional SFS (shape from shading) algorithms have been developed under the assumption of orthographic projection. However, the assumption is not valid when an object is not far away from the camera and, therefore, it causes severe reconstruction error in many real applications. In this research, we develop a new iterative algorithm for recovering surface heights from shaded images obtained with perspective projection. By dividing an image into a set of nonoverlapping triangular domains and approximating a smooth surface by the union of triangular surface patches, we can relate image brightness in the image plane directly to surface nodal heights in the world space via a linearized reflectance map based on the perspective projection model. To determine the surface height, we consider the minimization of a cost functional defined to be the sum of squares of the brightness error by solving a system of equations parameterized by nodal heights. Furthermore, we apply a successive linearization scheme in which the linearization of the reflectance map is performed with respect to surface nodal heights obtained from the previous iteration so that the approximation error of the reflectance map is reduced and accuracy of the reconstructed surface is improved iteratively. The proposed method reconstructs surface heights directly and does not require any additional integrability constraint. Simulation results for synthetic and real images are demonstrated to show the performance and efficiency of our new method.