Target tracking is currently a hot research topic in Computer Vision and has a wide range of use in many research fields. However, due to factors such as occlusion, fast motion, blur and scale variation, tracking meth...
Target tracking is currently a hot research topic in Computer Vision and has a wide range of use in many research fields. However, due to factors such as occlusion, fast motion, blur and scale variation, tracking method still needs to be deeply studied. In this paper, we propose a block target tracking method based on multi-convolutional layer features and Kernel correlation filter. Our method divides the tracking process into two parts: target position estimation and target scale estimation. First, we block the target frame based on the condition number. Second, we extract the features by the convolutional layer and apply it to the kernel correlation filter to get the center position of different block targets. With the reliability of different blocks measured by the Barker coefficient, the overall target position center is obtained. Then, the affine transformation is adopted to achieve the scale adaptation. The algorithm in this paper is evaluated by the public video sequences in OTB-2013. Numerous experimental results demonstrate that the proposed tracking method can achieve target scale adaptation and effectively improve the tracking accuracy.
A new image enhancement algorithm based on Retinex theory is proposed to solve the problem of bad visual effect of an image in low-light conditions. First, an image is converted from the RGB color space to the HSV col...
详细信息
A new image enhancement algorithm based on Retinex theory is proposed to solve the problem of bad visual effect of an image in low-light conditions. First, an image is converted from the RGB color space to the HSV color space to get the V channel. Next, the illuminations are respectively estimated by the guided filtering and the variational framework on the V channel and combined into a new illumination by average gradient. The new reflectance is calculated using V channel and the new illumination. Then a new V channel obtained by multiplying the new illumination and reflectance is processed with contrast limited adaptive histogram equalization(CLAHE). Finally, the new image in HSV space is converted back to RGB space to obtain the enhanced image. Experimental results show that the proposed method has better subjective quality and objective quality than existing methods.
Colourisation is a kind of computer-aided technology which automatically adds colours to greyscale images. This paper presents a scribble-based colourisation method which treats the flat and edge pixels differently. F...
详细信息
A siamese network tracking algorithm based on hierarchical attention mechanism is proposed in this paper. In order to obtain more robust target tracking results, different layer features are fused effectively. In the ...
A siamese network tracking algorithm based on hierarchical attention mechanism is proposed in this paper. In order to obtain more robust target tracking results, different layer features are fused effectively. In the process of extracting features, attention mechanism is used to recalibrate the feature map, and AdaBoost algorithm is used to weight the target feature map, which improves the reliability of the response map. Besides, the Inception module is also introduced which not only increases the width of the network and the adaptability of the siamese network to the scale, but also reduces the parameters and improves the speed of network training. Experimental results show that this method can effectively solve the impact of background clutter and improve the accuracy of tracking.
Voice conversion （VC） based on Gaussian mixture model （GMM） is the most classic and common method which converts the source spectrum to target spectrum. However this method is prone to over-fitting because of its ...
详细信息
Voice conversion （VC） based on Gaussian mixture model （GMM） is the most classic and common method which converts the source spectrum to target spectrum. However this method is prone to over-fitting because of its frame-by-frame conversion. The VC with non-negative matrix factorization （NMF） is presented in this paper, which can keep spectrum from over-fitting by adjusting the size of basis vector （dictionary）. In order to realize the non-linear mapping better, kernel NMF （KNMF） is adopted to achieve spectrum mapping. In addition, to increase the accuracy of conversion, KNMF combined with GMM （GKNMF） is also introduced into VC. In the end, KNMF, GKNMF, GMM, principal component regression （PCR）, PCR combined with GMM （GPCR）, partial least square regression （PLSR）, NMF correlation-based frequency warping （NMF-CFW） and deep neural network （DNN） methods are compared with each other. The proposed GKNMF gets better performance in both objective evaluation and subjective evaluation.
As a common malignant tumor disease, hepatocellular carcinoma is the most common cancers in the world. The incidence of hepatocellular carcinoma in China is higher than that in the world. Therefore, it is very importa...
As a common malignant tumor disease, hepatocellular carcinoma is the most common cancers in the world. The incidence of hepatocellular carcinoma in China is higher than that in the world. Therefore, it is very important for doctors to separate liver and tumor from CT images by means of computer-aided diagnosis and treatment. In this paper, a multiscale DC-CUNets network liver tumor segmentation method is proposed to enhance the fusion of multi-phase image features in CT, the scale of liver tumors, and the optimization of network training process. (1) A multistage CT liver tumor segmentation method based on two-channel cascaded U-Nets (DC-CUNets) is proposed. The liver was segmented using the first-order U-Net, and then the segmented area of interest of the liver was input into the second-order U-Net network to segment liver tumors. We designed two-channel U-Nets to learn the image characteristics of CT images in arterial and venous phases respectively, and to achieve two-channel feature fusion through feature cascade to improve the overall accuracy of liver tumor segmentation.(2) A multistage CT liver tumor segmentation method based on multiscale DC-CUNets was proposed. For the scale problem of liver tumors, we designed a two-layer multiscale void convolution module to obtain image features at different scales for large, medium and small tumors, and fuse the multiscale features at the output of the module. We have replaced the convolution layer of the fourth module in the second-order two-channel liver tumor segmentation U-Nets by the two-layer multiscale cavity convolution module to implement multiscale DC-CUNets.
Millimeter-wave(MMW) radar sensing is one of the most promising technologies to provide safe navigation for autonomous vehicles due to its expected high-resolution imaging capability However, driverless cars have high...
Millimeter-wave(MMW) radar sensing is one of the most promising technologies to provide safe navigation for autonomous vehicles due to its expected high-resolution imaging capability However, driverless cars have higher request for different environment and light conditions. Therefore, millimetre-wave imaging is of paramount importance for complex load scenario. In this paper, we have built models of pavement pits and bulges and analysed their with differences ways of antennas. A comparison of the imaging performance of experimental systems operating at a MMW radar and a Lidar is presented with the analysis of features for initial image interpretation Experimental images of the complex road surface are made by a 94GHz frequency-modulated continuous-wave (FMCW) radar technique with 3mm wavelength.
Although wavefront parallel processing(WPP) proposed in the HEVC standard and various inter frame WPP algorithms can achieved comparatively high parallelism, their scalab.lity for its parallelism is still very limited...
详细信息
Although wavefront parallel processing(WPP) proposed in the HEVC standard and various inter frame WPP algorithms can achieved comparatively high parallelism, their scalab.lity for its parallelism is still very limited due to various dependencies introduced in spatial and temporal prediction in HEVC. In this paper, through pixel correlation analysis, establishment of CTU node model, multi-core resource allocation strategy, we proposed an intra/inter-frame joint WPP coding algorithm for multi-core platform which can significantly improve the parallelism, while achieved good results in bit rate, PSNR, and acceleration ratio. Experiments on standard HD test sequence show that the proposed algorithm can lead to up to 2x, 1.3x speed up compared with the original WPP and IWF parallelism.
The performance of wavelet transform based image compressed sensing coding algorithms severely rely on the level of wavelet transform. To this end, this paper investigated the combination of mutil-level wavelet full s...
详细信息
The performance of wavelet transform based image compressed sensing coding algorithms severely rely on the level of wavelet transform. To this end, this paper investigated the combination of mutil-level wavelet full sub-band in compressed sensing framework. Firstly, we constructed a full sub-band coefficient sparse vector of mutil-level discrete wavelet transform. Secondly, we designed a weight matrix to improve measurement matrix. Finally, sparsity vector was processed by compressed sensing to get the measured value. Compared with the exiting algorithms, the experimental results of the proposed algorithm show that the PSNRs of reconstructed images is improve up to 1～2 dB under the same objective quality.
This paper presents a fast quality scalab.e video coding method based on compressed sensing(CS). The proposed method obtained the coding scheme of the enhancement MJU by using the interlayer and spatial correlation an...
详细信息
This paper presents a fast quality scalab.e video coding method based on compressed sensing(CS). The proposed method obtained the coding scheme of the enhancement MJU by using the interlayer and spatial correlation and kept the base layer's coding scheme unchanged. And the part in the enhancement layer which needed to be fine quantified was combined with the compressed sensing theory selectively which based on the sparsity of the signal and the complexity of the reconstruction. In order to satisfy the coding syntax structure of the reference software, the measurement value got by compressed sensing was complemented by 0 s and the flag bit was set to distinguish the special sub-blocks coded by CS. Experimental results show that the proposed algorithm can effectively improve the efficiency of scalab.e video coding and reduce the computational complexity.
暂无评论