最新刊期
卷 22 ， 期 3 ， 2017
- 摘要：In recent years, underwater image enhancement and restoration methods have become a crucial and difficult research focus because underwater images and videos play more and more important roles in the navy, marine environmental protection, and marine engineering. At present, there are relatively few available reviews or surveys on the research progress of underwater image enhancement and restoration methods. To obtain a more comprehensive understanding and promote the rapid progress of this research field, this survey systematically reviewed the research progress of underwater image enhancement and restoration methods. Based on numerous references, algorithms are classified and discussed on the basis of the use of a physical model. The principles, characteristics, and experimental methods of each algorithm are summarized. This survey analyzes the aspects of representative algorithms in detail. The evaluation system of underwater image quality is introduced. The representative methods are assessed with subjective and objective evaluation methods. The existing problems and developmental trend of this research field are highlighted. The causes of underwater image degradation, the principles of underwater image enhancement and restoration methods, and the development of the evaluation system of underwater image quality are summarized. Moreover, the existing problems and developmental trends of this research field are discussed. As an emerging research field, underwater image enhancement and restoration methods have potential research applications. However, the existing limitations of this research field require further research.关键词：underwater image;image degradation;underwater image enhancement;underwater image restoration;image quality assessment11369|935|45更新时间：2024-05-08
- 摘要：Numerous illegal pirating sites and programs have emerged with the rapid development of digital network technology because digital media can be easily copied and tampered with. Digital watermarking technology, which is an effective solution to image copyright protection, content authentication, integrity, and other issues, has become a crucial research topic in recent years. The singular values of images received via singular value decomposition (SVD) are strongly resistant to various attacks. Therefore, scholars have proposed various SVD-based watermarking methods. We proposed the novel boost normed singular value decomposition (BN-SVD) to solve the problems of diagonal distortion and false positives, which are caused by SVD. The method involves establishing a parameter to identify the optimum scaling factor for efficient robustness to improve existing SVD algorithms. Considering the inconsistencies between invisibility and robustness caused by embedding watermarks in original images by traditional watermarking algorithms, a novel zero-watermarking scheme based on BN-SVD is proposed. Zero-watermarking is defined by the registration and certification processes. In registration, an algorithm uses the major characteristics of a digital image to construct zero-watermarking registration information. The information is stored in a centralized authentication center. In certification, the watermark information is restored by using the digital images to be certified and the data stored in a centralized authentication center. First, a low-frequency approximation sub-graph was established from an original image that was decomposed by discrete wavelet transform (DWT) to non-overlapping image blocks. The low-frequency approximation sub-graph generated a low-frequency coefficient matrix via discrete cosine transform (DCT). BN-SVD was used to each block matrix to achieve a maximum singular value. A characteristic vector was created by comparing the maximum singular value with the average of the maximum singular value. The watermark image was disposed with Arnold transformation and Logistic map to obtain an encrypted and scrambling watermark image. Finally, the characteristic vector was set as an initial value and the scrambling encrypted watermark image as a controlling input value are both sent into a cellular neural network (CNN). By setting up a feedback template, the control template and threshold value of CNN determined the specific reversible logic operation. The output image was the registration image of zero-watermarking. The registration image was saved in the certification center to verify the copyright of the image. Experimental results indicated that all watermark images extracted by the proposed method do not exhibit diagonal marks when heterogenic attacks are imposed on original images. The proposed method overcame the diagonal distortion problem in the diagonal experiment. A one-to-one relation between the singular value vector and the image was established by introducing a parameter where different images have various singular value vectors. Therefore, the major characteristic of different images can be represented by different singular value vectors in a false-positive rate experiment. The normalized correlation between the extracted watermark image from BN-SVD and the original watermark image was below 50%. A low false-positive rate was observed. In the robustness experiment, we imposed various types of attacks, including JPEG compression, noise, filtering, rotating, and shear on the images. In the JPEG compression attack, the normalized correlation reached up to 99%. In the noise attack, the normalized correlation exceeded 97%. In the filter attack, the normalized correlation exceeded 98%. In the rotating attack, the normalized correlation was higher than 96%. In the shear attack, the normalized correlation exceeded 98%. These results indicated the need to improve high-shear attacks. A parameter will be used to modify the singular values of a matrix, which enhances the robustness of the algorithm and eliminates the false-positive and diagonal distortion problems of SVD. Introducing CNN is advantageous because the attacker cannot determine specific parameters. Therefore, the attacker extract the watermark image by reversible logic operation to improve the security of the watermarking. The parallel image processing of CNN can be achieved with hardware, which makes the algorithm applicable in occasions with higher real-time requirement.关键词：Arnold transform;boost normed singular value decomposition (BN-SVD);cellular neural network (CNN);discrete wavelet transform (DWT);discrete cosine transform (DCT);Logistic map;zero-watermarking2953|409|13更新时间：2024-05-08
- 摘要：Object tracking is an important research area in computer vision and has been widely adopted both in military and civilian applications. Improving tracking accuracy and stability in realistic scenarios that involve appearance change, occlusion, and illumination change is still difficult for practical application. A tracking method based on the phase congruency transformation and metric learning was presented to solve the aforementioned problem. This study formulates object tracking as a matching task to find a candidate, which is most similar to the target model, over the subsequent image frames. This process is largely controlled by two factors:the selected features that characterize objects and the distance metric used to determine the closest match in the selected feature space. First, the features were extracted by phase congruency transformation. Combining the advantages of ensemble learning and support vector machine (SVM), we then introduce a type of ensemble metric learning to obtain a distance metric matrix utilizing a small number of training data extracted from the fore sequence of images. Most approaches directly solve the optimal metric matrix and induce a large increase in the calculation as the feature dimension increases. In contrast, our method indirectly obtains the projection matrix by learning multiple projection vectors; thus, it is simple and efficient even with high-dimension features. Candidates are obtained by Markov chain Monte Carlo sampling and calculate the distance from the target model utilizing the learned metric matrix in the tracking process. The candidate with the smallest distance value is regarded as the target. Moreover, the object model and metric matrix are constantly updated with new training data extracted during tracking for adaptability. The effectiveness of the algorithm has been verified on several challenging video sequences that contain a dynamic background, appearance changes, and occlusions. The AEMTrack algorithm proposed in this study is clearly smaller on both the mean and standard deviation of the location error than those from three mainstream methods. Together with the quantitative assessment of tracking a successful rate, experimental results show that the accuracy of our proposed method even exceeds several mainstream methods in existing tracking studies and has appropriate stability. This study designs and realizes a new tracking method based on metric learning. A metric matrix is learned and tends to maximize the distance between samples of different classes using a small amount of training data sampled from an image sequence during tracking. The metric learning process is decomposed into multiple independent linear SVM, which can be executed in parallel implementation. This method can also result in dimension reduction; thus, it is efficient even in high-dimensional space. New targets and background samples are also applied to update the model in the tracking process; hence, the algorithm is adaptive. This tracking method has suitable generality because no limitation on the feature dimension of the target exists. Experimental results show that the proposed method can obtain an accurate and stable tracking effect in the complex scene, including appearance and illumination changes.关键词：visual tracking;phase congruency;metric learning;appearance change;adaptive2792|404|1更新时间：2024-05-08
- 摘要：Stereoscopic three-dimensional (3D) video services, which aim to provide realistic and immersive experiences, have gained considerable acceptance and interest. Visual saliency detection can automatically predict, locate, and identify important visual information, as well as help machines to effectively filter valuable information from high-volume multimedia data. Saliency detection models are widely studied for static or dynamic 2D scenes. However, the saliency problem of stereoscopic 3D videos has received less attention. Moreover, few studies are related to dynamic 3D scenes. Given that 3D characteristics, such as depth and visual fatigue, affect the visual attention of humans, the saliency models of static or dynamic 2D scenes are not directly applicable for 3D scenes. To address the gap in the literature, we propose a novel model for 3D salient region detection in stereoscopic videos. The model utilizes multi-dimensional, perceptual, and binocular characteristics. The proposed model computes the visual salient region for stereoscopic videos from spatial, depth, and temporal domains of stereoscopic videos. The proposed algorithm is partitioned into four blocks:the measures of spatial, depth, temporal (motion) saliency, and fusion of the three conspicuity maps. In the spatial saliency module, the algorithm considers the spatial saliency in each frame of videos as a visual attention dimension. The Bayesian probabilistic framework is adopted to calculate the 2D static conspicuity map. The spatial saliency in the framework emerges naturally as self-information of visual features. These visual features are obtained from the spatial natural statistics of each stereoscopic 3D video frame rather than from a single test frame. In the depth saliency module, the algorithm considers depth as an additional visual attention dimension. Depth signals have specific characteristics that differ from those of natural signals. Therefore, the measure of depth saliency is derived from depth-perception characteristics. The model extracts the foreground saliency from a disparity map, which is combined with depth contrast to generate a depth conspicuity map. In the motion (temporal) saliency module, the algorithm considers motion as another visual dimension. The optical flow algorithm is applied to acquire the inter-frame motion information between adjacent frames. To reduce the computational complexity of optical flow algorithms, the model first extracts the salient region of the current frame in accordance with the previously obtained spatial conspicuity map and depth conspicuity map. The Lucas-Kanade optical flow algorithm is adopted to calculate the motion characteristics between local salient regions of adjacent frames, and the motion conspicuity map is produced by the regional motion vector map. In the fusion step, a new pooling approach is developed to combine the three conspicuity maps to obtain the final saliency map for stereoscopic 3D videos. This fusion approach is based on the principle that human visual systems simultaneously focus on a unique salient region and divert attention to several salient regions in a saliency map. To generate the final saliency maps of stereoscopic videos, the proposed approach replaces the conventional average weighted sum for the fusion of different features and uses a fusion method that is based on global-local difference. We evaluated the proposed scheme for stereoscopic video sequences with various scenarios. Moreover, we compared the proposed model with five other state-of-the-art saliency detection models. The experimental results indicated that the proposed model is efficient, effective, and has superior precision and recall with an 80% precision and 72% recall rate. The proposed model demonstrated its efficiency and effectiveness in saliency detection for stereoscopic videos. The model can be applied to stereoscopic videos or image coding, stereoscopic videos or image quality assessment, and object detection and recognition.关键词：stereoscopic video;stereoscopic saliency detection;visual attention;binocular perceptual characteristics;depth saliency;motion saliency3078|454|3更新时间：2024-05-08
- 摘要：The saliency detection of RGB images has gained popularity in recent years. Saliency detection aims to quickly and accurately find salient objects in an image. Existing saliency detection methods utilize center prior, boundary prior, and color contrast methods. The center prior method is based on the assumption that salient objects are in the center of the image. The boundary prior method is based on the assumption that the image boundary is in the background instead of the foreground. The color contrast method is based on the principle that objects and their surroundings have varying contrast. Diffusion-based compactness method on the basis of the principle that salient objects typically have compact spatial distributions, whereas background regions have a wider distribution over the entire image. Fusion is required given that each method has its own advantages and weaknesses. In photography, important objects are placed at photographic composition intersections or arranged along photographic composition lines. Inspired by this photographic composition rule, we proposed saliency detection method based on the center rectangle composition prior, which assumes that the salient object is near the central rectangular composition lines. To supplement the weakness caused by assumptions- for example, the salient object is not near the center rectangular composition lines- we fuse compactness method to correct its result from spatial distribution. The compactness method calculates the spatial variance of each superpixel, as well as computes the spatial distance of the superpixels from the photographic composition intersection or the center of image. The choice of photographic composition intersection lies on cover range of the four intersections in initial saliency map. First, the image was converted into superpixels and a graph model was described. Then, we set the superpixels in the central rectangle composition lines as query nodes, extracted their color feature to rank with all the other regions by manifold ranking, and then generated the center rectangle composition lines saliency map. Third, we computed the spatial variance of each superpixel to generate compactness relation distribution saliency map. Fourth, we found all the photographic composition as intersections, which are salient in a center rectangle composition lines saliency map. Finally, we calculated and combined the spatial distance of superpixels from them to generate photographic composition intersection saliency map. If no singular data are found, the center point of image was selected. Finally, the final saliency map was obtained by fusing the center rectangle composition lines saliency map, center rectangle composition intersection saliency map, and compactness relationship distribution saliency map. Experiments were performed to compare the performances of nine different methods using MSRA-1000, CSSD, ECSSD, and THUS-10000 databases. Results showed that our final saliency maps were closer to the ground truth and our method had better precision-recall curve and higher F-measure value. Our method had an average running time of 0.673 s for one image, thus achieving the real-time requirement. Moreover, experiments showed that although Monica can produce the original saliency results generated by other methods, all of them were greatly improved to a similar accuracy level after optimization by center rectangle composition prior calculation. Our proposed method outperforms state-of-the-art methods given its effectiveness, robustness, and real-time application regardless of the complexity of the image or size of the salient object. These advantages are conferred by center rectangle composition prior and fusion of center rectangle composition lines prior, center rectangle composition intersection prior, and the compactness relationship.关键词：salient object detection;center rectangle composition prior;manifold ranking;center rectangle composition lines;center rectangle composition intersection;compactness relationship distribution3503|421|2更新时间：2024-05-08
- 摘要：The integral values of some successive equidistant subintervals are known in practical areas, whereas the usual function values at the knots are not given in interpolation problems. We propose a multilevel integro cubic spline quasi-interpolation for function approximation from given integral values over successive subintervals and multilevel spline quasi-interpolation. We used the linear combination of the given integral values to approximate function values at knots. The multilevel cubic spline quasi-interpolation operator was defined with the classical cubic spline quasi-interpolation and its corresponding error function. Finally, we obtained its polynomial reproducing property and error estimate. The proposed method, together with the existing integro cubic spline quasi-interpolation, was tested by two infinitely differentiable functions. Numerical experiments showed that the proposed method possessed better approximation behaviors and numerical convergence orders compared with the integro cubic spline quasi-interpolation. Multilevel integro cubic spline quasi-interpolation can successfully approximate the original function and its first and second-order derivative functions over the global interval. This process has good approximation behavior and numerical convergence compared with the existing integro spline quasi-interpolation. Moreover, the proposed method of function reconstruction from the integral values of successive subintervals is universally applicable.关键词：quasi-interpolation;multilevel;cubic B-splines;integral values;error analysis2534|353|2更新时间：2024-05-08
- 摘要：Various existing methods cannot reliably estimate the normal vectors for a point cloud model to smooth sharp features during point cloud processing. To address this problem, we developed a novel method based on Gaussian mapping to estimate the normal vectors of a scattered point cloud with sharp features. First, the normal vectors and feature points were roughly estimated by principal component analysis method. The feature points and their neighborhood points were mapped into a Gaussian sphere. Then, the K-means clustering algorithm was employed to segment data on the Gaussian sphere to several sub-clusters. Normal vector of a point is accurately estimated with the anisotropy neighborhood points that corresponded to the optimal sub-cluster to fit surface. Last, the effectiveness of the proposed method was validated by measuring the average deviation of the estimated normal vector from the standard normal vector. The estimated normal vectors were used in surface reconstruction to verify the feature-preserving property of the proposed method. Experimental results demonstrated that the least average deviation is close to zero. The method can accurately estimate the normal for noisy data. The reconstructed model maintains original geometry when the normal is used as input for the surface reconstruction algorithm. Compared with other normal estimation methods, the proposed method can more accurately estimate the normal vectors of points. The proposed method can accurately estimate the normal vector of a point model with sharp features. The method also exhibits high adaptability and robustness for point clouds with noise.关键词：normal estimation;reverse engineering;Gaussian mapping;principal component analysis;isotropy neighborhood;anisotropy neighborhood3183|472|4更新时间：2024-05-08
- 摘要：Synthetic aperture radar (SAR) image interpretation is an important aspect of SAR processing. Speckle noises proliferate in SAR images. Given that speckles cause problems with the interpretation of SAR images, speckles must be reduced to obtain high-quality images in SAR follow-up processing. Edge preservation is also a crucial aspect to consider. However, these tasks are inconsistent. Therefore, an efficient algorithm is needed to solve this problem. To achieve speckle reduction and preserve edges, we propose speckle reduction based on Euclidean distance anisotropic diffusion. The main model of the proposed method of speckle reduction is based on SRAD method. SRAD is a modified P-M method, which can reduce additive noise. However, speckle noise is multiplicative noise. The modified P-M method, SRAD, can reduce multiplicative noise. Thus, anisotropic diffusion is successfully applied in SAR image processing. The current study proposed anisotropic diffusion based on a novel edge-detection method. First, to maintain edge information, edges were detected, which can be performed by Euclidean distance. The value of the Euclidean distance was lower than the set threshold. The pixel points that were used to compare the two sides were considered non-edge areas. Otherwise, the points belonged to the edge area. The threshold was set by calculating the mean of all Euclidean distances. Second, a new anisotropic diffusion coefficient function was established based on the results of the first step of the study. The coefficient function determined the scale of diffusion. The established mathematical model calculated the diffusing capacity of all pixel points in SAR images. Finally, the model of anisotropic diffusion was developed by following the SRAD method. The calculated results can update the intensity value of all pixel points in SAR images. Anisotropic diffusion exhibited new behaviors because Euclidean distance was used in Euclidean distance anisotropic diffusion. The accurate calculations of the mean value and variance of speckle noises were difficult to obtain and significantly influence the result of speckle reduction. The proposed method can avoid the estimation of the mean value and variance. This paper uses several anisotropic diffusion methods on two TanDEM-X images. The result showed that these methods can effectively reduce speckle noises when images contained weak speckle noises. However, the proposed method yielded better results when images contained strong speckle noises. Euclidean distance anisotropic diffusion can effectively maintain edge, unlike other methods. A novel method to reduce speckles in SAR images is proposed. This method is categorized as anisotropic diffusion and is called Euclidean distance anisotropic diffusion. Euclidean distance anisotropic diffusion is a modified SRAD method based on Euclidean distance. It combines the SRAD model with Euclidean distance to effectively reduce speckles. The experimental result showed that the method can reduce speckle noise in areas with high and low concentrations of noise speckles.关键词：speckle;noise reduction;anisotropic diffusion;Euclidean distance;multiplicative noise2796|373|0更新时间：2024-05-08
- 摘要：Long-distance tracking is an important and challenging task in video surveillance. Existing tracking methods may fail when a target is completed occluded and is treated as a new target upon reappearance. Moreover, trackers are often confused by targets that appear similar. To address these problems, we propose a tracking algorithm that is aided by target recognition based on deep learning. The proposed method solves problems with tracking by identifying the corresponding relationship of objects detected between different frames. When an old target reappears, the algorithm can resume its tracking trajectory based on deep learning networks. Hence, the performance of tracking multiple and similar targets is improved. Experiments were conducted by comparing the standard dataset with other algorithms. Results showed that the proposed method can address occlusion, overlapping, and improve the performance of long-distance tracking. Therefore, the proposed method can continuously and effectively perform tracking. We propose a novel object tracking algorithm that is aided by recognition based on deep learning. The experimental results demonstrated the advantages of the proposed method in addressing the problem of a completely occluded object. Therefore, the proposed algorithm is suitable for the continuous tracking of multiple targets in monitoring videos.关键词：multi-target tracking;tracking based on recognition;deep learning;long distance tracking;trajectory recover7621|498|5更新时间：2024-05-08
- 摘要：The current cross-scenario clothing retrieval framework is based on the torso of the body. The retrieval results are refined by clothing recognition, which leads to lost clothing information and cross-scenario clothing recognition. We proposed a new clothing segmentation algorithm and clothing recognition method, which were based on domain-adaptive dictionary learning. First, we proposed an over-segmentation hierarchical fusion algorithm with pose estimation to segment intact clothing items and retrieve similar clothing images. During clothing recognition, the intermediate domain dictionaries between product clothing dataset and daily clothing dataset were sequentially learned to improve the accuracy of classifiers and the adaptability of the dictionary for clothing style recognition in different scenarios. To verify the efficiency of the proposed method, experiments were performed with Fashionista dataset, which a large public datasets, and our developed datasets. Experiments showed that the precision of the proposed method was 62.1% and 63.4% for the upper body and lower body, respectively, which indicated that the proposed method outperformed state-of-the-art methods in terms of clothing segmentation, clothing recognition, and clothing retrieval. To address the problem of current cross-scenario clothing retrieval, we proposed a novel hierarchical fusion clothing segmentation algorithm and domain-adaptive dictionary learning to recognize clothing attributes. The proposed method ensures the integrity of clothing and improves the precision of cross-scenario retrieval and style recognition.关键词：context based image retrieval;clothing retrieval;super-pixel;adaptive domain dictionary learning4135|382|3更新时间：2024-05-08
- 摘要：Images that are produced by polarization imaging have certain characteristics that reveal the advantages of fog goal. An effective quantitative evaluation method, however, has yet to be developed for polarization images. Given that the parameters of polarization images are generated by the original pattern, existing methods cannot effectively evaluate analytical images. The current study presents an evaluation method for the quality of polarization images under foggy conditions, as well as aims to compare the quality of images under different fog conditions. The relationship between the characteristics and the subjective observation of fog was verified from the view of image quality analysis. A polarization analysis of the characteristic parameters and factors of "analytical distortion sensitive" was conducted by analyzing the process of all polarization parameters and the influence of fog on image quality. These factors were based on the structural characteristics of the spatial statistical characteristics of natural scenes and images. Then, the corresponding Stokes parameters were introduced. A unified evaluation model was developed based on Mahalanobis distance. The experiment analyzed indoor simulated fog scene samples, simulated fog samples, and images under actual conditions. Experiments and the indoor simulation of a foggy environment were conducted with samples. Samples under foggy conditions were analyzed and evaluated. The validity and consistency of the subjective and objective experiments were determined with the three types of samples. The experimental results show that the CC values and RMS values of the map evaluated by the proposed algorithm are 0.930 2 and 4.593 2, respectively, and the CC value and RMS value of the map are respectively 0.877 1 and 0.995 0, algorithm has high accuracy. The SROCC value of is 0.939 0, the SROCC value of is 0.786 1, the objective score of the algorithm is consistent with the subjective score. The algorithm is better for the identification of the quality evolution relation of the polarization resolution parameter images under different fog conditions, The objective evaluation results are in accordance with the subjective analysis. In this paper, a comprehensive quality evaluation model based on polarization parameter image is proposed, which can accurately evaluate the and images in the parametric image by extracting the characteristic parameters and the Stokes parameters. The algorithm has high accuracy and good subjective and objective consistency, which can reflect the quality and correlation of the polarization parameter image, and can solve the problem of polarization imaging quality evaluation under the condition of fog.关键词：foggy image;polarization imaging;no-reference quality assessment;subjective and objective consistency3761|403|1更新时间：2024-05-08
- 摘要：Image inpainting is crucial for image processing. However, image inpainting methods produce restored images with unsatisfactory textures. Therefore, to effectively maintain image textures, we propose two image inpainting algorithms based on continued fraction interpolation. The proposed algorithms are based on continued fraction interpolation. The intensity of a damaged point is interpolated from the information of the surrounding pixel points. The two proposed interpolation methods are based on different interpolation functions and interpolation windows to repair different types of scratching texture images:the inpainting algorithm based, which is based on Thiele interpolation, and the inpainting algorithm, which is based on Newton-Thiele interpolation. Moreover, we propose the solutions to singular point and translation problems in interpolation. To demonstrate the superiority of the proposed algorithms, several experiments were conducted with scratching images. Subjective and objective evaluations were employed. The objective evaluation compared the peak signal-to-noise ratio (PSNR) and running time among algorithms. The experimental results showed that the proposed algorithms exhibited better visual effect, higher PSNR, and shorter running time than those of current popular inpainting algorithms. The PSNR of the proposed algorithm was 44.79 dB, and its running time was 0.53 s. The proposed inpainting algorithm, which is based on Thiele interpolation, is more suitable for scratching images with perpendicular textures. By contrast, the inpainting algorithm, which is based on Newton-Thiele interpolation, is more appropriate for complex texture images.关键词：image inpainting;scratch;continued fractions;interpolation;Newton-Thiele;Thiele4144|364|1更新时间：2024-05-08
- 摘要：Modern medical imaging devices, such as CT and MRI, can provide detailed data representations of the interior of the human body. Realistic volume rendering incorporates Monte Carlo path-tracing algorithm, which addresses the visualization of these data to provide photorealistic images. However, Monte Carlo path-tracing volume rendering demands high storage volume and computational power. Therefore, it can only be interactively visualized on clusters with installed high-performance GPU. Current mobile devices, such as smart phones and tablets, significantly contribute to mobile computing. The realistic volume rendering of high-volume datasets at interactive update rates is still computationally expensive for current hand-held devices because of the limitations of computational capability and power consumption. Remote rendering helps address this issue by running heavy-volume rendering computation on a powerful GPU-accelerated server and transmitting the rendering results to the mobile client for display and interaction. A volume renderer that utilizes Monte Carlo path-tracing (MCPT) techniques was set up to progressively generate realistic images. A dedicated web server sent rendering results to and received user inputs from mobile devices. Renderer servers can produce rendered images at a high updating rate, whereas web server streams render images at a slower rate than a renderer server, particularly when network conditions are poor. Low rendering efficiency is caused by unbalanced data transmission and processing rate between renderer servers and web servers. This paper introduced optimized coupling and performance between a graphic server and web server to improve rendering efficiency and maintain system interactivity. The coupling algorithm can balance the output speed of renderer servers and transfer speed of web servers by adaptively adjusting the rendering iterations of each output image and keeping renderer servers busy before completing rendering tasks. To directly compare the performance of the coupling algorithm with that of a connected volume renderer server with web server, we set up an experiment that recorded the total frames received by browsers and the duration of the entire rendering process after user interaction was terminated. The experiment utilized Manix, Mecanix, VisMale, and Bonsai volume datasets. Experimental results indicated that using the coupling algorithm between renderer servers and web servers significantly reduced the time required to achieve the final high-quality image, unlike the direct connection between renderer servers and web servers. The results also revealed that the renderer server connects with web servers via the coupling algorithm, which allows end users to obtain the final high quality image in nearly the same amount of time, regardless of network condition. Web-based realistic remote volume rendering system is proposed to provide end users with access to high-performance rendering services using any HTML5 supported device. The coupling algorithm that links renderers and Web servers can adjust the output speed of the renderer depending on the overall capacity of a network while maintaining the performance of the renderer. Thus, end users quickly obtain the final rendered images.关键词：Web;remote rendering;realistic;image transmission;interaction optimization4758|655|1更新时间：2024-05-08
- 摘要：Existing 3D handheld scanners are usually integrated with image sensors that can directly generate texture images. However, the resolution of texture images is usually low, and problems, such as highlights, shadows, and shading, exist. This paper proposed a texture reconstruction method based on multiple photos. First, the image and geometric model are registered via feature-matching method. Second, the direct and accurate correspondence between image pixels and texture pixels is established by adopting a special encoding location texture. Then, the diffuse component is solved with simultaneous equations, which are established by location texture. Finally, we obtain intrinsic textures via the improved fusion method on the basis of hybrid weight. The proposed method was used to reconstruct the intrinsic texture of the three experimental models. The method was compared with a 3D scanning device to generate texture. Photos were directly utilized to generate texture. Results showed that the method is simple to operate, easy to use, and can produce clear and intrinsic textures without highlights and shadings. Experimental results showed that the reconstruction of texture quality in resolution, color reduction, and consistency is better than that of the original texture. The method has high accuracy and robustness and can meet the demand of high-quality texture reconstruction.关键词：texture reconstruction;texture mapping;texture blending;location mapping;location texture3898|386|2更新时间：2024-05-08
- 摘要：The increasing demand for personalized pictures has encouraged the fusion of images and texts to form a new type of media, which has resulted in the new field of image processing:visual media automatic design. Visual automatic design focuses on technology that can automatically add aesthetic text watermarks to users' pictures. In this study, we optimized the location and color schemes of text watermarks based on the principle of design and computer vision. We selected the best location of text watermarks on the basis of visual saliency theory and composition theory. We selected a suitable color was by considering visual contrast and color harmonization. To successfully combine texts and images, we adaptively matched the original images with the template of the color wheel to select the appropriate text color. Comparing the digital watermarking of Tencent (73.25%:9.32%) and of the pictures in National Geographic (Chinese version) (97.8%:2.8%) with that of the proposed algorithm showed that the algorithm automatically provided satisfactory results. The proposed algorithm improves the design and enhances the aesthetics of the combined text and image layouts whether the current watermark camera or the simple editing is involved.关键词：automatic layout;color harmonization;visual saliency;image composition;visual contrast5171|402|0更新时间：2024-05-08
- Postal code：100190
- Tel：010-58887035/58887030/58887418 Email：jig@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备05080539号-4
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0