SOM And SVM Based Feature Extraction And Detection Of Objects In Images
Abstract— Image retrieval is a technique in which the image is retrieved using feature extraction or tagging so that it can be
used for a variety of applications. Although there are various techniques implemented for the image retrieval such as using
Fuzzy Classifier, the methodology used for the image retrieval using Fuzzy is efficient but has more error rate and image
retrieval rate is low. Hence, an efficient technique is proposed here for the image retrieval using SOM based clustering, so
that the clustering of the specified objects in the image is done and image is retrieved by training those features of the image.
The methodology adopted here provides more retrieval rate as compared to the previous technique of image retrieval. The
idea is to convert the image into HSV color space providing multi-resolution structure and applying SOM clustering to
cluster and detect the features of the image, now segment only the features, which contains object in the image.