Denoising and Motion Artifact Removal Using Deformable Kernel Prediction Neural Network for Color-Intensified CMOS.
Zhenghao HanLi LiWeiqi JinXia WangGangcheng JiaoXuan LiuHailin WangPublished in: Sensors (2021)
Keyphrases
- denoising
- neural network
- prediction model
- color images
- image denoising
- prediction accuracy
- motion analysis
- image sequences
- motion prediction
- color information
- total variation
- power consumption
- motion model
- multi layer perceptron
- translation invariant
- color space
- prediction error
- high speed
- optical flow
- artificial neural networks
- neural network model
- partial least squares regression
- deformable surface
- beltrami flow
- elman network
- image processing
- human motion
- motion estimation
- low cost
- kernel function
- image sharpening
- back propagation
- multiple cues
- support vector
- moving objects
- chaotic time series
- articulated objects
- radial basis function
- noisy images
- natural images
- denoising methods
- wavelet packet
- deformable models