VFBGM: Very Fast Block-Wise Gradient Magnitude-Based Quality Metric for Infrared Images

Infrared images are visual representations generated by processing infrared radiation data captured through infrared thermal imaging cameras. Characterized by strong penetration, immunity to visible light interference, and high concealment, infrared images play a vital role in military, ecological conservation, industrial, and other fields. However, during the generation, acquisition, transmission, and storage of infrared images, their quality is often degraded by factors such as non-uniform noise, Gaussian noise, and compression distortions, posing challenges to image processing and subsequent applications. To address this issue, this paper proposes a novel infrared image quality assessment (IQA) model, termed very fast block-wise gradient magnitude (VFBGM), which can serve as an accelerated optimizer to enhance infrared image quality and ensure target detection accuracy. The VFBGM quantifies image quality by calculating the mean gradient magnitude between the original and distorted images in a block-wise manner. Experiments conducted on the Infrared Image Quality Evaluation Database (I2QED) demonstrate that our proposed VFBGM achieves exceptional performance, significantly outperforming the majority of mainstream and state-of-the-art methods in terms of efficacy, while also delivering the fastest computational speed.