A Novel Facial Image Resolution Enhancement Algorithm using Pixels Homogeneity

Vikram Mutneja, Dr. Satvir Singh

Abstract


Facial image processing is an important step in the surveillance cameras based biometric identification systems. The resolution of the facial images from surveillance videos is usually very less owing to the factors such as hardware constraints, distance between the subjects and CCTVs etc. In this work, a novel spatial domain based face resolution enhancement technique using single LR image has been proposed. It is based on the homogeneity levels of the pixels in respective neighboring four quadrants in 3×3 scanning window, used as LR patch which is replaced by interpolated 5×5 HR patch. The kernel matrix is generated for all the pixels in the source image which encodes the homogeneity of all pixels in image. The performance analysis of proposed technique has been done for different values of resolution enhancement factors, in terms of quality metrics Structural Similarity Index Measure (SSIM), Peak Signal to Noise Ratio (PSNR), Signal to Noise Ratio (SNR) and Mean Square Error (MSE). From the results, we contend that the proposed algorithm is very effective in generating good quality HR facial image from single LR facial image. Keywords: Face Resolution Enhancement, Low Resolution Surveillance Videos, Spatial Domain Processing, Neighbourhood Processing, Image Quality Enhancement

Full Text:

PDF


DOI: https://doi.org/10.26483/ijarcs.v8i3.3046

Refbacks

  • There are currently no refbacks.




Copyright (c) 2017 International Journal of Advanced Research in Computer Science