Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Do you really want to delete this prezi?
Neither you, nor the coeditors you shared it with will be able to recover it again.
Make your likes visible on Facebook?
You can change this under Settings & Account at any time.
Content Based Image Retrieval
Transcript of Content Based Image Retrieval
What it is
Difference from Text Retrieval
Problems with Image Retrieval
Types of Image Retrieval
Conclusion Text retrieval – simple, don’t need any sort of intricate system to identify words Image retrieval – complex, how do you identify a cat from a bed? Problems with Image Retrieval Images are not text! Ideas for how you search images? You search for a black car. Text-wise, that’s pretty simple, but how does that work image-wise? So you find a bunch of cars, but what if a car is under enough shade that it seems black? Types Keywords Denotes prominent pieces of the image
E.g. computer, CD, mouse , phone Description Tag that has a brief overview of what is in the image
E.g. Two soccer teams chase after the ball. One team is wearing red while the other is white. It is a sunny day. They are in a stadium Advantages and Disadvantages Advantages Easy to conceptualize Everything is manually done No errors due to machine User-friendly, but not developer-friendly Disadvantages Resources needed to implement are very high-end What if each picture needs over 200 words to fully describe it? Given a reference database of unlabeled images, retrieve images similar to a new query image based only on visual content. Content - Based Image Retrieval Methods Local Binary Patterns (LBP) Colors and Texture Reference Database of Mammographies Image Regions as Query Examples Local Derivative Patterns (LDP) Locality-Sensitive Hashing Self-Supporting Retrieval Map Algorithm User Query Image Image Database Image Feature Extraction Retrieved Image Distance Measure Features Color Shape They used gradient vector flow (GVF) fields to obtain the edge image. Algorithm: (edge image computation)
1. Read the image and convert it to gray scale.
2. Blur the image using a Gaussian filter.
3. Compute the gradient map of the blurred image.
4. Compute GVF. (100 iterations and = 0.2 )
5. Filter out only strong edge responses using k s , where s is the standard deviation of the GVF. (k – value used is 2.5).
6. Converge onto edge pixels satisfying the force balance condition yielding edge image. They have designed an algorithm for finding the minimum cost matching based on most similar highest priority (MSHP) principle using the adjacency matrix of the bipartite graph. The integrated minimum cost match distance between images is now defined as: Integrated image matching Content Based Image Retrieval based on Color, Texture and Shape features using Image and its complement Jagadeesh Pujari and P. S. Hiremath EXPERIMENTAL RESULTS Texture