CODEWITHSUNDEEP
image-processingcomputer-visionfeature-detectionfeature-descriptorglcm
Akshay Pai
Akshay Pai

Reputation: 11

Performace comparison between GLCM and LIRE for image matching

I want to create an image matching application. When I searched for it, I found two good approaches.

1) LIRE - Lucene Image Retrieval can be used to perform the image matching , where it implements various image matching algorithms and and also indexes feature descriptors which can be used later to be matched against the query image.

2) GLCM - Gray Scale Co-occurance Matrix, where the features of a gray scale image is stored in the form of a matrix called the Co-occurrence matrix, and when an image has to be matched, then the matrix of the given image is calculated and important features are matched to calculate the similarity.

Both these approaches seem scalable and fast. But I am not able to decide which might perform better. So any help regarding this or the situations during which they are an advantage is what I am looking for.

Upvotes: 0

Views: 255

Answers (1)

Akshay Pai
Akshay Pai

Reputation: 11

After a lot of research, I have found the difference between these two.

LIRe is a library that provides multiple implementations of various image matching algorithms that work on local or global features of an image. Also, LIRe provides indexing and retrieval solutions off the box that makes image matching and result retrieval very efficient and fast.

On th other hand, GLCM is an algorithm that works upon Gray scale images and tries finding how similar the images are. It just an algorithm where as LIRe is a whole library of several fast implemented algorithms.

LIRe does not support GLCM but the ones that LIRe support are very fast. The book on LIRe has spoken in great detail and has helped me a lot in Image matching and retrieval.

Upvotes: 0

Related Questions