The fracturing process of granite cubes (400 × 400 × 400 mm3) subjected to blast loading was investigated using data from Digital image Correlation (DIC) analysis on images captured with a high-speed camera....
详细信息
Heatmap regression has been used for landmark localization for quite a while now. Most of the methods use a very deep stack of bottleneck modules for heatmap classification stage, followed by heatmap regression to ext...
详细信息
ISBN:
(纸本)9781538664209
Heatmap regression has been used for landmark localization for quite a while now. Most of the methods use a very deep stack of bottleneck modules for heatmap classification stage, followed by heatmap regression to extract the keypoints. In this paper, we present a single dendritic CNN, termed as Pose Conditioned Dendritic Convolution Neural Network (PCD-CNN), where a classification network is followed by a second and modular classification network, trained in an end to end fashion to obtain accurate landmark points. Following a Bayesian formulation, we disentangle the 3D pose of a face image explicitly by conditioning the landmark estimation on pose, making it different from multi-tasking approaches. Extensive experimentation shows that conditioning on pose reduces the localization error by making it agnostic to face pose. The proposed model can be extended to yield variable number of landmark points and hence broadening its applicability to other datasets. Instead of increasing depth or width of the network, we train the CNN efficiently with Mask-Softmax Loss and hard sample mining to achieve upto 15% reduction in error compared to state-of-the-art methods for extreme and medium pose face images from challenging datasets including AFLW, AFW, COFW and IBUG.
LBP coefficients are essential and determine the priority of gray differences. The objectives of this paper are to reveal this and propose a method for finding an optimal priority through the genetic algorithm. On the...
详细信息
ISBN:
(纸本)9781509064540
LBP coefficients are essential and determine the priority of gray differences. The objectives of this paper are to reveal this and propose a method for finding an optimal priority through the genetic algorithm. On the other hand, the genetic operators such as initialization and cross-over operators, generate invalid coefficients, defective chromosomes. This paper also recommends a rectifying method for correcting defective chromosomes. Results on the FERET and Extended Yale B datasets indicate that the proposed method has markedly higher recognition rates than LBP.
image segmentation is one of the most important steps in analysis of features in image data. In order to acquire stable image of blood vessel, frame-to-frame matching as a step in preprocessing has be made in solving ...
详细信息
In this paper a new method based on Region Growing (RG) and Spectral Cluster (SC) for segmentation of synthetic aperture radar (SAR) images is introduced. In the proposed method first RG is applied to the SAR images i...
详细信息
ISBN:
(纸本)9781509064540
In this paper a new method based on Region Growing (RG) and Spectral Cluster (SC) for segmentation of synthetic aperture radar (SAR) images is introduced. In the proposed method first RG is applied to the SAR images in order to find the edge and then segmentation is done using SC method. The proposed method (RG+SC) is compared with some state-of-the-art segmentation algorithms on real SAR image. Obtained results show the efficiency of the proposed approach.
Aqueous phase liquid (APLs) leakage and spillage into the subsurface system, leading to groundwater contamination is an issue that needs to be addressed. This paper aims to investigate the APLs migration characteristi...
详细信息
Aqueous phase liquid (APLs) leakage and spillage into the subsurface system, leading to groundwater contamination is an issue that needs to be addressed. This paper aims to investigate the APLs migration characteristics in fractured non-isothermal double-porosity soil. A laboratory experiment was conducted to observe and monitor the characteristics of the soil structure and APLs migration in heated deformable double-porosity soil using digital image processing technique. The results show rapid liquid migration for the fractured soil samples. The time taken for the liquid to migrate under the application of heat is less for sample with low moisture content due to faster dry off and rapid evaporation. It can be concluded that APLs migration under vibration and non-isothermal effect is highly influenced by the soil sample structure, the soil fractured pattern, the soil water content, and the applied heat in the soil.
This paper proposes a multi-objective approach to improving the component substitution (CS) based pansharpening method by obtaining the adaptive weights. The non-dominated sorting genetic algorithm II (NSGA-II) is emp...
详细信息
ISBN:
(纸本)9781509064540
This paper proposes a multi-objective approach to improving the component substitution (CS) based pansharpening method by obtaining the adaptive weights. The non-dominated sorting genetic algorithm II (NSGA-II) is employed to simultaneously optimize two objective functions. The inverse of the Correlation Coefficient (CC) and a weighted sum of the Erreur Relative Globale Adimensionnelle de Synthese (ERGAS) in the spectral and spatial domains are used as the objective functions. The use of a multi-objective approach in the CS technique allows optimizing the fused image in terms of both spatial and spectral resolutions. Simulation results show that the proposed method outperforms popular CS-based fusion methods.
In this paper, we propose an algorithm for tracking of moving objects in video sequences. Our method uses Kalman filter to predict the location of target and exploits superpixel based tracking algorithm to find the re...
详细信息
ISBN:
(纸本)9781509064540
In this paper, we propose an algorithm for tracking of moving objects in video sequences. Our method uses Kalman filter to predict the location of target and exploits superpixel based tracking algorithm to find the real position of target in a search region surrounding the predicted location. The motion dynamics and equations from mechanics physics are used to design a Kalman filter with assumption of constant acceleration motion. Using this Kalman filter makes our method able to handle long lasting occlusions. We have also devised a scheme that helps the tracker to find the target after long lasting occlusions.
The proceedings contain 125 papers. The topics discussed include: two-stream federated learning: reduce the communication costs;a new update strategy for blocks with low correlation in 3-D recursive search;eye movemen...
ISBN:
(纸本)9781538644584
The proceedings contain 125 papers. The topics discussed include: two-stream federated learning: reduce the communication costs;a new update strategy for blocks with low correlation in 3-D recursive search;eye movement pattern modeling and visual comfort viewing S3D images;motion trajectory based spatial-temporal degradation measurement for video quality assessment;two-pass rate control for constant quality in high efficiency video coding;adaptive motion vector prediction for omnidirectional video;generative adversarial network-based frame extrapolation for video coding;a CNN-based in-loop filter with CU classification for HEVC;synthesizing 3D acoustic-articulatory mapping trajectories: predicting articulatory movements by long-term recurrent convolutional neural network;analysis of smoothed LHE methods for processing images with optical illusions;and deep network with spatial and channel attention for person re-identification.
Nowadays, image processing is getting more popular due to the daily increase of diverse data acquisition methods such as digital scanners and cameras. Due to the high volume of archived documents, automatic document c...
详细信息
ISBN:
(纸本)9781509064540
Nowadays, image processing is getting more popular due to the daily increase of diverse data acquisition methods such as digital scanners and cameras. Due to the high volume of archived documents, automatic document classification methods can help to save the time and space in digital document organization. Logos in official and business documents are used to identify document identities. Different approaches have been used for logo recognition yet, many of which has complex computations to achieve a high level of precision. In this paper, a novel algorithm for accurate logo recognition with low level of computational complexity is proposed based on Local Binary pattern (LBP). We proposed PerLogo dataset consisting 850 images of 10 different classes of logos has been proposed in this paper. Through 3 separate experiments over 50, 60, 70 images per each class the proposed system has been evaluated. Experimental results show that recognition rate is increased with increasing the number of training images per class. Experimental results show the recognition accuracy of 98% when 0.09 salt and pepper noise are added to the test images, which is more than 95% accuracy proposed by the state-of-the-art approaches achieving 95% accuracy.
暂无评论