Image Querying with Complex Wavelets: Results of Experiments1.32003/12/11 01:29:46 US/Central2003/12/16 02:04:23.441 US/CentralTomMowadtm@rice.eduVenkatChandrasekaranvenkatc@rice.eduTomMowadtm@rice.eduVenkatChandrasekaranvenkatc@rice.eduimage queryingdiscrete wavelet transformcomplex discrete wavelet transformGives quantiative results for the different approaches, highlighting strenghts and weaknesses of the old and new approaches.
We decided that for testing purposes we would test images from three categories:
hand-drawn images (using mspaint)blurred images (which emulates scanned images)shifted images
From each category, we performed five tests using query images based on images from our database. For every test image we used, except for hand drawn image 5 (shown below), the expected image was returned in the top three percent of matched images when using the complex wavelet algorithm. Results using the real wavelet algorithm were similar, except for on shifted images, where the query images had significantly lower rankings (on the range of top 15%) than from our algorithm.
hand drawn 1actual imagehand drawn 2actual imagehand drawn 3actual imagehand drawn 4actual imagehand drawn 5actual imageblurred 1actual imageblurred 2actual imageblurred 3actual imageblurred 4actual imageblurred 5actual imageshifted 1actual imageshifted 2actual imageshifted 3actual imageshifted 4actual imageshifted 5actual image