Otomotivde güvenli video iletimi

Video monitoring became essential for the ongoing development of car systems, delivering instantaneous visual feedback for driver assistance, object detection, and safety features. As these video streams were used more extensively, the danger of security breaches with the potential to compromise the integrity and authenticity of visual data also increased. To offset such risks, digital watermarking emerged as a tool, with the capability of embedding hidden authentication data in video streams without affecting perceptual quality. This work evaluated the performance of four watermarking schemes—Base Effec- tive Algorithm, Sequenced Watermarking, Multi-Subband Watermarking, and HMAC- Enhanced Watermarking—under varying attack scenarios, including additive Gaussian noise, compression, cropping, and image swapping. Using objective quality metrics such as Peak Signal-to-Noise Ratio (PSNR) and Normalized Cross-Correlation (NCC), we quantified visual quality versus watermark robustness trade-offs. Experiments were conducted under both light and heavy Gaussian noise scenarios to simulate real-world tampering. It was concluded that under low attack levels, visual quality was best preserved by the Base Algorithm, whereas Multi-Subband Watermarking performed better under more aggressive manipulation conditions. Sequenced Watermarking proved promising indetectingframe-levelforgeries, andHMAC-basedembeddingprovidedcryptographic security. These findings provided a strategic basis for selecting watermarking methods based on security requirements for video surveillance in automotive systems.