ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

Message (Text) Extraction System from Noisy-Image Regions Using Morphological Dialation Technique in Forensics Investigation

Journal: International Journal of Computer Science and Mobile Computing - IJCSMC (Vol.9, No. 12)

Publication Date:

Authors : ; ; ; ; ;

Page : 95-102

Keywords : Horizontal and vertical projection; Morphological dilation; Geometric properties of text; Image binirization and Gap filling method;

Source : Downloadexternal Find it from : Google Scholarexternal


In cyber/forensics investigation, it was observed that there is a need to extract text/messages from scene images especially from a noisy-image zones/regions, hence the attention was drawn to study text detection and recognition in natural images using computer vision systems. Text detection and recognition in natural images has applications in computer vision systems like image retrieval, registration plate number detection, automatic traffic sign detection, could be of help for visually impaired people. Scene text, however, has complicated background, blur image, partly occluded text, variations in font-styles, image noise and ranging illumination. Hence scene text recognition could be a difficult computer vision problem. In the previous works, paper connected component method is used to extract the text from background. In this work, horizontal and vertical projection profiles, geometric properties of text, image binirization and gap filling method has been used to extract the text from scene images while the morphological dilation method was introduce to separate text background of the images. precision rate and recall rate of the proposed method and that of the other existing methods were compared and discovered that the performance of our proposed method is excellent overall according to the result of the metrics variables. Therefore, the proposed method is proved to be efficient for extracting the text/messages from the outdoor scene images. Finally text is extracted from images for optimal result.

Last modified: 2021-04-01 03:58:49