Dehaze-AI-
🌫️ Haze Removal Using Dark Prior Channel 🌫️
📚 Project Overview
Welcome to our final year project! We’re addressing the common issue of haziness in images with the Dark Prior Channel method. 🌫️ Dust, haze, and fog can obscure details and diminish image quality, making it hard to see what’s important. Our project aims to tackle this by refining and enhancing ground truth images, removing unwanted distortions for clearer and more vibrant visuals. 🚀✨
Example: Hilly Valley
Here’s an example demonstrating our method on a hilly valley scene:
🔍 Methodology
The Dark Prior Channel technique is designed to improve image clarity by leveraging the unique properties of haze-free images. Here’s how it works:
- Pixel Intensity Analysis: In images without haze, some pixels exhibit very low intensity in at least one color channel. 🌈
- Haze Estimation: By identifying these dark pixels, we estimate the haze in the image. 📉
- Haze Removal: We then apply our findings to clear the haze, dust, and fog, enhancing the overall image quality. 🖼️
🚧 Challenges
Despite achieving promising results, we face several challenges:
- Resource Limitations: Our ability to perform high-level haze removal is constrained by limited resources. 🛠️💡
- Complexity of Haze: Some images present more complex haze patterns that are harder to remove completely. 🌀
🌟 Achievements
- Enhanced Image Clarity: Our method has successfully improved the visibility of key details in hazy images. 🏆
- Valuable Insights: The project contributes to a better understanding of haze removal techniques and their practical applications. 📊🔍
📸 Visual Results
Here are some examples of our work:
- Before Dehazing:
- After Dehazing:
🛠️ How to Use
To see our method in action or integrate it into your projects, check out our code and examples provided in this repository. For detailed instructions and usage, refer to the documentation.
Thank you for exploring our project! Feel free to provide feedback or contribute. 🙌💬