iarjetIARJETIARJET2708-5163https://doi.org/10.47310/iarjet.2025.v06i01.004Grey Scale Image DeNoising Using Adaptive Threshold in Co-sine DomainImage De-Noising is essential for images transmitted through a noisy communications channel. Several approaches have been presented to address this scenario in different communications mediums and for different noise types. In this work, a de-noising scheme is proposed in which an adaptive algorithm is utilized to retain certain Cosine coefficients. The selection of these coefficients, based on a data-driven threshold, makes the resultant image less noisy for Gaussian noise. The adaptive algorithm decompose the noisy image into Cosine and Haar domains and the decomposition process halts when a predefined limit is reached. The proposed scheme is tested with grey scale images that are publicly available and compared with the state-of-the-art systems. Peak Signal to Noise Ratio (PSNR) is used as a metric besides the human visual inspection. As shown in the results, the proposed system performs better than the system under comparison in terms of the used metrics.