Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Scene Text Segmentation by Applying Trimmed Median Filter using Energetic Edge


Affiliations
1 Department of Computer Applications, Manomaniam Sundaranar University, India
2 Department of Computer Applications, Noorul Islam University, India
     

   Subscribe/Renew Journal


This proposed method is an accurate and strong method for detecting texts in natural scene images. There are many cases that text regions are not clearly noTable.by background regions due to brightness or illumination variations. The proposed scene text finding process finds out the scene text contents from the natural scene images using the sophisticated edge revealing methods, the local directional number feature and linked map generation process. This proposed method takes natural scene image as input and detects the scene text regions. The detected scene text regions are marked for visual identification for human eyes.

Keywords

Noise Reduction, Energetic Edge Detection, Local Directional Number, Linked Map, Non Seen Text Rejection.
Subscription Login to verify subscription
User
Notifications
Font Size


  • Scene Text Segmentation by Applying Trimmed Median Filter using Energetic Edge

Abstract Views: 396  |  PDF Views: 6

Authors

T. Beula Bell
Department of Computer Applications, Manomaniam Sundaranar University, India
M. K. Jeya Kumar
Department of Computer Applications, Noorul Islam University, India

Abstract


This proposed method is an accurate and strong method for detecting texts in natural scene images. There are many cases that text regions are not clearly noTable.by background regions due to brightness or illumination variations. The proposed scene text finding process finds out the scene text contents from the natural scene images using the sophisticated edge revealing methods, the local directional number feature and linked map generation process. This proposed method takes natural scene image as input and detects the scene text regions. The detected scene text regions are marked for visual identification for human eyes.

Keywords


Noise Reduction, Energetic Edge Detection, Local Directional Number, Linked Map, Non Seen Text Rejection.

References