Development of a Modified Predictive Coding Algorithm for High-Resolution Image Compression in Real-Time Structural Health Monitoring of Metallic Component

Abstract

Real-time monitoring requirements of structural integrity in metallic components across aerospace and automotive and civil engineering sectors drive the need to create reliable image compression methods. The detection of small defects including cracks and corrosion as well as deformations depends greatly on high-resolution imaging systems. The massive amount of generated image data creates problems with both data storage and its transmission requirements. The research advocates for developing a Modified Predictive Coding (MPC) algorithm which specializes in real-time high-resolution image compression for Structural Health Monitoring (SHM) systems. The proposed Modified Predictive Coding algorithm builds on predictive coding yet adds adaptive context modeling tools alongside edge-preserving processes to balance visual quality with high compression ratio fulfillment. This method surpasses traditional approaches by making prediction parameter modifications which happen automatically according to image local features in order to protect essential diagnostic information. Testing of the algorithm took place using a high-resolution image dataset containing metallic surface images which faced different stress scenarios. The proposed method underwent performance evaluations that applied Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM) and Compression Ratio (CR) metrics for assessment. The experimental findings show that MPC algorithm achieves higher compression efficiency and maintains visual quality better than JPEG2000 and SPIHT methods do. Its fast processing abilities along with low latency system make MPC strongly compatible with deploying SHM systems utilizing edge devices and IoT methodologies. The research analyzes how the algorithm would function with sensor networks and cloud-based analytics systems for improving predictive maintenance decision capabilities. The Modified Predictive Coding algorithm proves to be a suitable technology for efficient high-quality image compression which enables prompt accurate assessment of metallic structures in critical infrastructure. This research initiative creates an opportunity to study advanced compression methods which unite machine learning with predictive coding for smart SHM system applications.