Image Retrieval Using Feature Extraction

Download Document/Synopsis
Here we proposed a system where similar images are fetched based on image features. As images are more suspected to noise we used many preprocessing steps to remove noise . Image features are extracted using feature extraction method and these features are stored into database. Query image feature values are compared with other images in directory and similar image is extracted and displayed to user. Image is quantized in hsv color space into equal bins and the features extracted from hsv color. Filter is used to remove noise from image. Image is given as an input to the application, system find its nearest neighbor from the training set and system fetches nearest image to the input test image. System sorts images according to smallest distance. User must select an image and system will extract image based on query image features and will display similar image to user.


Advantages
  • This system works with any jpg images.
  • System finds out similar images which has similar query image feature value.
Disadvantages
  • System fetches only one image.
  • System works only with jpg image files.

-->