@article{Divakar_Singh_Bajpai_Kumar_2022, title={Image pattern recognition by edge detection using discrete wavelet transforms}, volume={2}, url={https://icda.rabek.org/index.php/about/article/view/3}, DOI={10.31181/jdaic10029042022k}, abstractNote={<p>Edge is the high-frequency part of an image and represents location where abrupt change takes place in the intensity of luminescence. Edge detection is the basic step of the feature extraction and pattern recognition of any image. Wavelet transforms extract low and high-frequency information of any signal separately. In two-dimensional wavelet transforms, an image is decomposed into four sub-images the one approximation image and three different images (horizontal, vertical, and diagonal images) in each decomposition level. The differences’ images show how the neighboring pixels differ in the horizontal, vertical and diagonal directions. The approximation coefficients are forced to zero and differences’ coefficients are inverse wavelet transformed. As the reconstructed image shows the edges of the image and describes its pattern. Using Haar wavelet at decomposition level 1, 2 and 3, the image pattern recognition by edge detection is performed and discussed.</p>
<p><em> </em></p>}, number={1}, journal={Journal of Decision Analytics and Intelligent Computing}, author={Divakar, Ravikant and Singh, Bijendra and Bajpai, Ashish and Kumar, Anil}, year={2022}, month={Apr.}, pages={26–35} }