I have a image set and all images has similar background except the object in the image if any. Is there any techniques in computer vision to compress all images by using the similarity feature?
Thomas Varghese I think what you're looking for is background subtraction. https://docs.opencv.org/3.4/d1/dc5/tutorial_background_subtraction.html
You could also try a learning based segmentation approach if you're familiar with DL techniques. Let me know if you have further questions.
Dear Thomas Varghese ,
Compression-based similarity measures employ in an unusual way general off-the-shelf compressors, by exploiting them to estimate the amount of information shared by any two objects. Such techniques, of which the most well-known is the Normalized Compression Distance (NCD).
https://core.ac.uk/download/pdf/30997967.pdf
Regards,
Shafagat
Yes, there are several techniques in computer vision that can be used to compress images based on their similarity features. One such technique is called "content-based image retrieval" (CBIR), which involves extracting features from images (such as color, texture, and shape) and using these features to retrieve similar images from a database.
Another technique that can be used for image compression is "image clustering", which involves grouping similar images together and then compressing them as a group. This can be done using algorithms such as K-means clustering or hierarchical clustering.
In addition, there are also "image similarity search" algorithms that can be used to find the most similar images to a given query image, which can be useful for compressing images that are very similar to each other.
Overall, there are many different techniques in computer vision that can be used to compress images based on their similarity features, and the best approach will depend on the specific application and requirements of the task.
- Badges
- Science topic