Object Recognition and Clustering based on Latent Semantic Analysis (LSA)

Abstract

Object Recognition and clustering are prime techniques in Computer Vision, Pattern Recognition, Artificial Intelligence and Robotics. Conventionally these techniques are implemented in Visual-Feature based methods. However, these methods have drawbacks they do not efficiently deal with the differences in shapes and colours of objects. Another method which uses semantic similarity to solve this kind of problem, i.e. Cosine Similarity method, but this method also has problems. The problems are synonymies and polysemies. In this paper we propose a method in which objects with different shapes and different colours which function similarly can be recognized and clustered. If the text printed on the object the semantic feature of that object is extracted and clustered according to semantic feature. Proposed method is based on semantic information so we conduct an experiment with the dataset of images which contains the packing cases of commercial products (e.g. Mobile, Laptop etc). Semantic information in dataset is retrieved using text extraction module and then the results of text extraction are passed through an Internet search module. Finally objects are described and clustered using the latent semantic analysis (LSA) module. The clustering results are more accurate than the Visual feature based method and cosine similarity based methods.
DOI: 10.17762/ijritcc2321-8169.1505120