Content-Based Image Retrieval (CBIR) is a technique for searching and retrieving images from a database based on their visual content, like color, texture, and shape, rather than metadata or keywords.
- Custom model implementation
- VGG model implementation
- Relevance Feedback
- Web based Application
- Basic UI with Bootstrap
- Readme
- Documents
- User ability to choose model
- Migrate to a real host
- Image upload and processing
- Retrieval of similar images using CBIR techniques
- Relevance feedback for refining results
- Responsive design with Bootstrap for seamless user experience
Image upload and processing Retrieval of similar images using CBIR techniques Relevance feedback for refining results Responsive design with Bootstrap for seamless user experience
Means converting images into numerical data that represent key visual properties.
Used histograms, color moments in `LAB` color space.
Used `Gabor` filters, Local Binary Patterns (LBP), or wavelet transforms.
Used edge detection algorithms like `Canny`, or `Sobel`.
Scale features to ensure consistency.
Fuse different features into a single vector if needed.
in CBIR is an interactive process where users mark retrieved images as relevant or not. The system then updates and refines search results based on this feedback, improving accuracy in finding similar images.
The VGG model is a type of deep learning network used for image recognition. It uses many small 3x3 filters stacked on top of each other to find patterns in images. The most common versions are VGG16 and VGG19, with 16 and 19 layers. It’s popular because it works well for tasks like classifying and comparing images.
Django is a Python web framework that simplifies building secure, scalable applications with features like ORM and authentication. Bootstrap is a front-end framework for creating responsive, visually appealing interfaces. Together, they streamline backend development and UI design.