Multi-Exposure Fusion with Guidance Information: Night Colour Image Enhancement for Roadside Units
Mogili Ravindar
Associate Professor
Dept. of Computer Science and Engineering
Jyothishmathi Institute of Technology and Science (JNTUH)
Karimnagar, Telangana, India
mogili.ravinder@jits.ac.in
Manchikatla Srivally
UG Student
Dept. of Computer Science and Engineering
Jyothishmathi Institute of Technology and Science (JNTUH)
Karimnagar, Telangana, India
srivallymanchikatla@gmail.com
Ch. Keerthi
Assistant professor
Dept. of Computer Science and Engineering
Jyothishmathi Institute of Technology and Science (JNTUH)
Karimnagar, Telangana, India
keerthi.chinthala@jits.ac.in
Mohammad Abid Pasha
UG Student
Dept. of Computer Science and Engineering
Jyothishmathi Institute of Technology and Science (JNTUH)
Karimnagar, Telangana, India
abidedunet@gmail.com
Vaishnavi Maddela
UG Student
Dept. of Computer Science and Engineering
Jyothishmathi Institute of Technology and Science (JNTUH)
Karimnagar, Telanagana, India
vaishnavi.maddela04@gmail.com
Khaja Bilaluddeen
UG Student
Dept. of Computer Science and Engineering
Jyothishmathi Institute of Technology and Science (JNTUH)
Karimnagar, Telangana, India
knoxbilaluddeen@gmail.com
Abstract—Night-time monitoring for roadside units is challenging due to poor illumination, low contrast, noise, and loss of important visual details such as traffic markings, obstacles, and vehicle registration numbers. Single-exposure images captured at night often fail to represent all scene information clearly, either losing details in dark regions or becoming overexposed in bright areas. To address this issue, this project proposes a Multi-Exposure Fusion with Guidance Information approach for night-time color image enhancement.
The proposed system takes multiple images of the same roadside scene captured at different exposure levels (dark, normal, and bright) as input. A guided fusion strategy is applied to combine the most informative regions from each exposure while preserving color consistency, edge details, and structural information. Image quality measures such as contrast, saturation, and edge strength are used to guide the fusion process. The final enhanced image provides improved visibility of road markings, vehicles, obstacles, and surrounding infrastructure under low-light conditions.
Keywords—Multi-Exposure Fusion, Night Image Enhancement, Guided Image Filtering, Roadside Units, Intelligent Transportation Systems, Low-Light Imaging
