| 引用本文: | 龙古灿,张小虎,于起峰.用于视频图像帧间运动补偿的深度卷积神经网络.[J].国防科技大学学报,2016,38(5):143-148.[点击复制] |
| LONG Gucan,ZHANG Xiaohu,YU Qifeng.Deep convolutional neural network for motion compensated frame interpolation[J].Journal of National University of Defense Technology,2016,38(5):143-148[点击复制] |
|
| |
|
|
| 本文已被：浏览 8168次 下载 7703次 |
| 用于视频图像帧间运动补偿的深度卷积神经网络 |
| 龙古灿1,2, 张小虎1,2, 于起峰1,2 |
|
(1.国防科技大学 航天科学与工程学院， 湖南 长沙 410073;2.
2.国防科技大学 湖南省图像测量与视觉导航重点实验室， 湖南 长沙 410073)
|
| 摘要: |
| 为探索深度学习理论在视频图像帧间运动补偿问题中的应用，提出一种用于视频图像帧间运动补偿的深度卷积神经网络。该网络由卷积模块和反卷积模块构成，可以处理不同分辨率输入图像并具备保持较完整图像细节的能力。利用具有时序一致性的视频图像序列构造训练样本，采用随机梯度下降法对设计的深度卷积神经网络进行训练。视觉效果和数值评估实验表明，训练得到的网络较传统方法能更有效地进行视频图像帧间运动补偿。 |
| 关键词: 深度学习 卷积神经网络 时序一致性 运动补偿帧插值 |
| DOI：10.11887/j.cn.201605022 |
| 投稿日期：2016-04-27 |
| 基金项目:国家重点基础研究发展计划资助项目（2013CB733100） |
|
| Deep convolutional neural network for motion compensated frame interpolation |
| LONG Gucan1,2, ZHANG Xiaohu1,2, YU Qifeng1,2 |
|
(1. College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China;2.
2. Hunan Key Laboratory of Videometrics and Vision Navigation, National University of Defense Technology, Changsha 410073, China)
|
| Abstract: |
| In order to explore the application of deep learning theory in the problem of motion compensated frame interpolation, a DCNN (deep convolutional neural network) built with convolutional blocks and deconvolutional blocks was proposed. The proposed DCNN is capable of processing input images with different resolutions and preserving fine grained image details. The temporal coherent image sequences were used to construct the training sample and the stochastic gradient descent method was adopted to train the designed DCNN. Qualitative and quantitative experiments show that the trained DCNN obtains better interpolated images than the traditional approach in two testing images sequences. |
| Keywords: deep learning convolutional neural network temporal coherence motion compensated frame interpolation |
|
|
|
|
|
