Abstract
AbstractIn this paper, the K-means algorithm for information extraction in the form of text from complex video images using 2D wavelet is presented. Haar wavelet and K-means algorithm-based simple hybrid approach are proposed in this paper. Haar wavelet is used to efficiently convert grayscale image to edge image. K-means algorithm is used for the information localization and segmentation process to split the background pixels from the text images. Large non-text background interrupts to extract the text information from any video images, so that non-text background can be identified and eliminated from the text images with the help of the hybrid approach. This algorithm is evaluated on the complex background video frames. The recall rate and the precision rate are obtained 99.01 and 95.75% of the proposed algorithm in video images.KeywordsText extractionVideo images2D Haar waveletK-means algorithm
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
