An Efficient and Fast Method for Image Super Resolution by Block Based Image up Sampling
Bhupesh Gupta, Gagandeep Singh, Mohit Chhabra, Amit Kumar Bindal, Avinash Sharma · 2023
In today’s environment image super resolution is the challenging task for researchers. This research work proposes an innovative upsampling technique for images in the domain of Discrete Cosine Transform, employing Orthogonal Rotation Invariant Moments (ORIMs). The method capitalizes on the inherent sparsity of DCT coefficients, effectively reconstructing high-resolution images from their low-resolution counterparts. By harnessing the unique properties of ORIMs, precise estimation of high-frequency DCT coefficients is achieved, ensuring superior image quality with minimal artifacts. Additionally, a strategic quantization scheme is implemented to optimize computational efficiency and memory usage. Extensive experimental evaluations demonstrate the method's superiority over conventional spatial domain interpolation and up-to-the-minute DCT domain procedures in terms of visual fidelity and Peak Signal-to-Noise Ratio. This DCT-based approach using ORIMs holds great promise for high-quality image upsampling in diverse applications.