Aerial images are densely captured with high resolution. It is an important research topic to efficiently process the aerial images. Benefited from its powerful computation power, GPU acceleration techniques have been...
详细信息
ISBN:
(纸本)9781538619377
Aerial images are densely captured with high resolution. It is an important research topic to efficiently process the aerial images. Benefited from its powerful computation power, GPU acceleration techniques have been widely used in imageprocessing, and many traditional techniques achieved real-time processing through using the GPU acceleration. In this research, we study the aerial image blending problems using GPU acceleration. We parallelized the Multi-band blending algorithm and achieved 30 times acceleration compared with CPU implementation.
This paper proposes an improved threshold segmentation to turn the image of fluffs into binary image, the noises and impurities in the image are eliminated by using Gauss filtering and the method of features statistic...
详细信息
ISBN:
(纸本)9781538619377
This paper proposes an improved threshold segmentation to turn the image of fluffs into binary image, the noises and impurities in the image are eliminated by using Gauss filtering and the method of features statistics of color fluffs, then the centroid coordinate of each color fluff is calculated and the location information of each color fluff is obtained, finally every color fluff is detected and eliminated. The experiment proved that the threshold segmentation has the characteristics of high detection rate and fast calculate speed.
This study aims to understand the characteristics of art movements through interdisciplinary research methods that include art history, network analysis, and imageprocessing. We conducted research at three levels-mac...
详细信息
ISBN:
(纸本)9781538619377
This study aims to understand the characteristics of art movements through interdisciplinary research methods that include art history, network analysis, and imageprocessing. We conducted research at three levels-macroscopic, mesoscopic, and microscopic-to reconsider the historical relevance of the art movement. In this study, we mainly focused on the difference between Dutch art and Impressionism, and the usage of blue pigment as a criterion to compare traditional and modern art groups. The results of the analyses provide a holistic perspective for enriching historical knowledge of art movements.
Brain tumor segmentation on MR images has significant clinical meaning due to glioblastomas which are the most lethal form of these tumors. Compared to manual segmentation, automatic segmentation system is superior in...
详细信息
Brain tumor segmentation on MR images has significant clinical meaning due to glioblastomas which are the most lethal form of these tumors. Compared to manual segmentation, automatic segmentation system is superior in timesaving and experience-insensitivity for doctors during clinical practice. However, its inherent contradiction is not addressed yet. i.e. imbalance of multi-class of different brain tissues. As such, we proposed a multi-class focal loss to make the loss function emphasis on bad-classified voxels in MR images. Our experiments based on the 3D UNet model proved that this method can significantly improve labeling and segmentation accuracy as compared to other loss layers.
Due to "semantic gap", the problem of image emotional semantic annotation has not been solved. In this parper, a method of emotion semantic annotation for cheongsam images based on Fusion Features has been p...
详细信息
ISBN:
(纸本)9781538619377
Due to "semantic gap", the problem of image emotional semantic annotation has not been solved. In this parper, a method of emotion semantic annotation for cheongsam images based on Fusion Features has been proposed. Multi-features including the color and texture are used to describe the content of the image. Then least squares support vector machine for regression which is optimized by particle swarm optimization is used to build the mapping between the feature space and emotional space. The experiment indicates that this method achieves good effect.
Regularization methods are effective in to solving ill-posed vision problems such as image denoising and restoration. These methods typically involve a smoothness/regularization term (prior) and a data term (fidelity)...
详细信息
An algorithm to detect car distance from a pair of stereo images is presented. It is useful for drivers to avoid collisions and ensure safety to keep the car at a constant distance from the car ahead. The conventional...
详细信息
ISBN:
(纸本)9781538619377
An algorithm to detect car distance from a pair of stereo images is presented. It is useful for drivers to avoid collisions and ensure safety to keep the car at a constant distance from the car ahead. The conventional distance detection method is based on image matching;the proposed algorithm is based on key-point matching. Key points are extracted from a stereo image pair by using Speeded Up Robust Features (SURF). The distance is calculated from 3D binocular disparity, the difference of position at the object.
Aiming at the needs of applications of home-based care for the aged and communities, this paper presents a seven-lead electrocardiograph (ECG) monitoring system based on Wi-Fi short-range wireless transmission technol...
详细信息
Aiming at the needs of applications of home-based care for the aged and communities, this paper presents a seven-lead electrocardiograph (ECG) monitoring system based on Wi-Fi short-range wireless transmission technology. The system mainly consists of two parts: the seven-lead ECG monitoring device and the ECG monitoring platform. The Seven-lead ECG monitoring device mainly consists of the ECG pre-processing module, integrated analog front end, the micro-controller, the Wi-Fi wireless transmission module and the power module. This system has the advantages of miniaturization, low power consumption and high performance. The ECG monitoring system uses Wi-Fi wireless transmission technology to send the ECG data of monitored by the seven-lead ECG monitoring device to the ECG monitoring platform of the monitoring center for real-time display of ECG waveforms. Doctors can obtain the health status of the patients who recuperates at home or the aged of community by the community ECG monitoring system and which can also remind the aged of abnormal ECG signals in time to go for a further physical examination for reducing the sudden death rate of the cardiovascular disease.
We design a bi-objective particle swarm optimization algorithm-based protein structure refinement method and the basic motivation is trying to solve the bias problem caused by minimizing only a single energy function ...
详细信息
ISBN:
(纸本)9781538619377
We design a bi-objective particle swarm optimization algorithm-based protein structure refinement method and the basic motivation is trying to solve the bias problem caused by minimizing only a single energy function due to the complexity and the diversity of different protein structures. The bi-objective energy function optimization strategy designed in our protocol extends the 1-D optimization to a new 2-D optimization driven by the artificial intelligence PSO engine.
Cryo-EM reconstruction technology is developing rapidly, while the 3D density map's resolution determination method is still unsatisfactory. In this paper, we propose a new multi-scale spectral signal to noise rat...
详细信息
ISBN:
(纸本)9781538619377
Cryo-EM reconstruction technology is developing rapidly, while the 3D density map's resolution determination method is still unsatisfactory. In this paper, we propose a new multi-scale spectral signal to noise ratio analysis method to determine a reasonable global resolution. We apply the mask idea and generate multi-scale masks to estimate the final resolution.
暂无评论