This paper proposes a novel geodesic saliency propagation method where detected salient objects may be isolated from both the background and other clutter by adding global considerations in the detection process. The ...
详细信息
ISBN:
(纸本)9781479923427
This paper proposes a novel geodesic saliency propagation method where detected salient objects may be isolated from both the background and other clutter by adding global considerations in the detection process. The method transmits saliency energy from a coarse saliency map to all image parts rather than from image boundaries in conventional cases. The coarse saliency map is computed using the combination of global contrast and Harris convex hull. Superpixels from pre-segmented image are used as pre-processing to further enhance the efficiency. The proposed propagation is geodesic distance assisted and retains the local connectivity of objects. It is capable of rendering a uniform saliency map while suppressing the background, leading to salient objects being popped out. Experiments were conducted on a benchmark dataset, visual comparisons and performance evaluations with 9 existing methods have shown that the proposed method is robust and achieves the state-of-the-art performance.
An effective approach for coverage holes was proposed, called Distance-assistant Coverage Hole Parching (DACHP) strategy. DACHP strategy would achieve the purpose of patching coverage holes using redundant sensor node...
详细信息
ISBN:
(纸本)9781629932170
An effective approach for coverage holes was proposed, called Distance-assistant Coverage Hole Parching (DACHP) strategy. DACHP strategy would achieve the purpose of patching coverage holes using redundant sensor nodes within the wireless sensor networks. Simulation results show, compared with DFNFP and random algorithm, for the same size and shape coverage hole, DACHP can awaken five redundant nodes to patch 92.19% of hole, yet DFNFP needs awaken nine redundant nodes and just patch 90.28%. It can be seen that DFNFP not only awaken four extra nodes, but also is lower 1.19% in patching rate of hole than DACHP. In addition, DACHP is more stable in utilization of redundant nodes and does not produce additional cost.
Foreground and background are treated without distinction at classification stage in most background subtraction algorithms. However, correct classification of foreground is the primary requirement, and thus misclassi...
详细信息
ISBN:
(纸本)9781479923427
Foreground and background are treated without distinction at classification stage in most background subtraction algorithms. However, correct classification of foreground is the primary requirement, and thus misclassification costs of the t-wo classes should be different. Based on this fact, we present a new method to introduce cost sensitivity into background subtraction, where a cost matrix is created to represent the costs of misclassification. Some items in the cost matrix are not constants, but functions of foreground occurence at each pixel location. By the use of such non-constant costs, detection rate of foreground is improved while increase of false alarms is prevented at the same time. Experiments demonstrate the effectiveness of the proposed algorithm.
We introduce a semi-automatic tracking method that can be utilized for the analysis of facial markers in the medical condition of facial palsy. Tracking of markers will help medical physicians in evaluating this medic...
详细信息
image registration is widely used in remote sensing imageprocessing. On one hand, non-subsamded Contourlet transform (NSCT) has the advantage of decomposing image in a flexible way;on the other hand, cross-cumulative...
详细信息
image registration is widely used in remote sensing imageprocessing. On one hand, non-subsamded Contourlet transform (NSCT) has the advantage of decomposing image in a flexible way;on the other hand, cross-cumulative residual entropy (CCRE) is effective in remote sensing image registration. Considering that, we propose a multimodal remote sensing image registration method which is based on cross-cumulative residual entropy and NSCT algorithm. First, the reference image and target image are decomposed with NSCT to obtain low frequency images, and then the cross-cumulative residual entropy of the obtained low frequency images is calculated. Set the cross-cumulative residual entropy as a similarity measurement. Secondly, Newton's method is employed to gain optimal parameters of the affine transformation model. Finally, the image registration is obtained with the optimal parameters. To validate our algorithm, we test two remote sensing images with our method. Simulation results show that the proposed method is able to find the global optimum rapidly and prevent dropping into a local minimum. In general, it is not only a fast and effective multimodal remote sensing image registration algorithm but also the one with high registration accuracy.
Remote sensing image with the characteristics of overall dim brightness, low-contrast and no obvious distinction between the target and background, makes the remote sensing image enhancement technology play an active ...
详细信息
Remote sensing image with the characteristics of overall dim brightness, low-contrast and no obvious distinction between the target and background, makes the remote sensing image enhancement technology play an active role for improving the contrast of the image and highlighting some local details. At present, the image multi-scale system has gained successful applications in imageprocessing, wavelet transform, Curverlet transform, Contourlet transform and some improved algorithms based on them. One of the most common shortcomings of the frameworks of the system Curverlet and Contourlet is lacking of providing a unified treatment between the continuity and the digital world. Shearlet system is the only one which satisfies this property, yet still optimally delivers sparse approximations of images. Based on Shearlet transform, we propose a new algorithm for remote sensing image enhancement. First, the remote sensing images are transformed into the low frequency coefficients and high frequent coefficients by Shearlet transform;second, we use fuzzy contrast enhancement on the low frequency coefficients of Shearlet transform;third, we use fuzzy enhancement on the high frequency coefficients of each scale and direction. Experimental results show that in the subjective side, we obtain a good visual effect, and objectively, the image entropy and mean have a significant promotion.
Orientation field is one of the intrinsic charac-teristics of fingerprints, and it plays very important role in many processing phases of fingerprint recognition such as enhancement, minutiae extraction and matching. ...
详细信息
ISBN:
(纸本)9781467321969
Orientation field is one of the intrinsic charac-teristics of fingerprints, and it plays very important role in many processing phases of fingerprint recognition such as enhancement, minutiae extraction and matching. Therefore, accurate estimation of orientation filed is much necessary. In this paper, a novel PDE-based method is proposed for regularization of orientation field for low-quality fingerprint images. The method consists of four steps. Firstly, the coarse orientation field is computed using traditional gradient-based approach. Secondly, the reliability map of the orientation field is computed based on a procedure of multiscale coherence analysis. Then the orientation in the low-reliable region is reconstructed with the surrounding data by means of image inpainting technique. Finally, nonlinear diffusion filtering with adaptive diffusivity is performed on the whole orientation field. Experiments on the NIST SD4 fingerprint database indicated that the proposed algorithm is capable to estimate the orien-tation field accurately, especially for poor-quality fingerprints, and it can be integrated into fingerprint recognition systems to improve the performance.
Non-photorealistic rendering (NPR) such as stylization has received much concern in area of computer graphics and imageprocessing. In this paper, a simple and effective method is proposed for streamline stylization. ...
详细信息
We study the task of interactive semantic labeling of a segmentation hierarchy. To this end we propose a framework interleaving two components: an automatic labeling step, based on a Conditional Random Field whose dep...
详细信息
暂无评论