Multi-Exposure Image Fusion Algorithm Evaluation: A Comparative Study Using HDR-CAVE and Expo-database

This paper presents a perceptual evaluation of multi-exposure image fusion algorithms utilizing two prominent datasets: HDR-CAVE and Expo-database.

1. The HDR-CAVE dataset comprises 60 high dynamic range images, each exhibiting 24-bit true color. These images capture diverse scenes, including indoor, outdoor, and natural landscapes. HDR-CAVE serves as a widely employed dataset for evaluating multi-exposure image fusion algorithms.

2. The Expo-database dataset encompasses 20 multi-exposure image sequences, with each sequence consisting of 3 images captured at varying exposure times. These images also represent diverse scenes such as indoor, outdoor, and architectural settings. Expo-database is a newer dataset commonly used for evaluating multi-exposure image fusion algorithms.