Enhance HDR Image Fusion with Algorithms(2 images)

I'm working on merging high-gain and low-gain grayscale images to create HDR images using the trans_mef algorithm. However, I believe the current model mainly focuses on preserving structure and features, while contrast and structural details play a crucial role in the fusion process, not just brightness.

As the trans_mef algorithm is typically designed to merge more than 2 images, we cannot modify it. Instead, we can either combine it with another algorithm or add more layers to achieve superior HDR effects. The goal is to enhance the evaluation indicators:
- MEF-SSIM should exceed 0.9
- MSE
- PSNR should surpass 25
- SSIM

The ultimate output should be an HDR image with a good contrast balance, capable of revealing details in both bright and dark areas.

I'm currently using the Python Imaging Library (PIL) for image processing. I would like to integrate Exposure Fusion with trans_mef for improved HDR results.

Specifically, I am looking for enhancements in:
- Edge preservation
- Textural details
- Brightness contrast

The ideal candidate for this project should have:
- Extensive experience with Python and PIL
- Strong understanding of image processing algorithms, particularly HDR techniques
- Ability to improve contrast and structural details in images.

Please note that the dataset is already available for this project.