Resolution Reconstruction Using Patch-Match And Wavelet Decomposition
Sangeetha Muthiah, A. Senthilrajan · 2022 6th International Conference on Intelligent Computing and Control Systems (ICICCS) · 2022
This paper proposes a new method to preserve edges and enhance image details while reconstructing image resolution based on Discrete Wavelet Transform (DWT) image decomposition combined with a patch-matching technique using a single low-resolution image. Discrete wavelet transformation on the low-resolution input provides LL, LH, HL, and HH subbands considered for further process. Later, bicubic interpolation approximates the three subbands LH, HL, and HL of the inputted low-resolution image. Next, the patch-match approach is employed to estimate the HF details of the input image that replaces the LL subband. Then, the inverse DWT is applied on the four subbands to obtain a sharper reconstructed image. The findings obtained on various images indicate that the proposed approach preserves edges and enhances the fine details with minimal distortion and smoothing. The quantitative and visual results of the presented method demonstrate its efficacy compared to existing techniques.