About this document
Image Retrieval Similarity Measurement Evaluation by ahsan is a document available to read on EtoBox.
The document evaluates different similarity measures that can be used for content-based image retrieval where images are represented as feature vectors. It describes commonly used distance measures like Minkowski distance (with variations like City Block and Euclidean), Cosine distance, χ2 statistics, Histogram intersection, and Mahalanobis distance. An experiment was conducted to evaluate these measures on a shape image database to determine the most accurate and efficient measure for shape-based image ret
- Author
- ahsan
- Language
- EN