Sarcouncil Journal of Engineering and Computer Sciences

Abstract: Wireless Image Transmission Systems (WITS) are highly susceptible to noise distortions, leading to degraded image quality and increased transmission errors. This study evaluates the effectiveness of Digital Signal Processing (DSP) algorithms, Kalman filtering, wavelet-based denoising, Wiener filtering, and median filtering for noise reduction in WITS under varying Signal-to-Noise Ratio (SNR) levels (5 dB to 20 dB). Performance is assessed using Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Bit Error Rate (BER). The results indicate that Kalman filtering achieves the highest PSNR (33.2 dB) and SSIM (0.96), along with the lowest BER (0.005) at SNR = 20 dB, making it the most effective method for noise suppression. Wavelet-based denoising emerges as a computationally efficient alternative, offering a balance between image quality and processing speed. Statistical analyses, including ANOVA and paired t-tests, confirm significant differences (p < 0.001) in performance among the algorithms. The findings suggest that while Kalman filtering provides superior noise reduction, wavelet-based denoising is more suitable for real-time applications. Future research should explore hybrid DSP techniques and deep learning-based models for enhanced noise suppression in wireless imaging