Resolution Enhancement in Video Using Super Interpolation Algorithm
S. Muthuselvan1, S. Rajaprakash2, J Hemanth Sai3, Byrapaneni Nikhil4
1S. Muthuselvan, Department of CSE, Aarupadai Veedu Institute of Technology, Chennai, Tamil Nadu.
2S. Rajaprakash, Department of CSE, Aarupadai Veedu Institute of Technology, Chennai, Tamil Nadu.
3Byrapaneni Nikhil, Department of CSE, Aarupadai Veedu Institute of Technology, Chennai, Tamil Nadu.
4J Hemanth Sai, Department of CSE, Aarupadai Veedu Institute of Technology, Chennai, Tamil Nadu.
Manuscript received on 02 June 2019 | Revised Manuscript received on 10 June 2019 | Manuscript published on 30 June 2019 | PP: 258-264 | Volume-8 Issue-8, June 2019 | Retrieval Number: H6329068819/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Super resolution technique in digital image and video frame for enhancing the imaging system to acquire high resolution from low resolution image/frame. The challenges in super resolution are video/image registration, Denoising, light variation, blur identification, computation efficiency and performance limits. In order to overcome the problem of high frequency details lost during reconstruction and computational complexity due to iterative methods in the existing super resolution techniques, this paper propose a work of fiction algorithm named Super Interpolation(SI) to achieve the low complex upscaling of video frames with High Resolution(HR). SI method consists of two phases: Upscaling Phase and Training Phase. In the training phase, a large set of external training images/video frames undergoes edge orientation analysis. The primary, upscaling phase, the LR video frame is upsampled and interpolated by bicubic interpolation method. Then the interpolated frame is subjected to edge detection by canny edge detector for frame smoothing. Frame sharpening is by local laplacian filter with edge preservation technique to get the reconstructed HR video frame. The proposed method tested on two different data sets YT dataset and Real Time (RT) – India dataset. It is found from the experimental results that the proposed method performs better than existing FRESH algorithm from both subjective visual effect and objective measurements of PSNR and SSIM with less computational time.
Keyword: Edge orientation, Gradient, Image upsampling, Resolution enhancement, Super interpolation.
Scope of the Article: Parallel and Distributed Algorithms