[Paper Review] Segmentation of Camera Captured Business Card Images for Mobile Devices

Ayatullah Faruk Mollah, Subhadip Basu+1 more|arXiv (Cornell University)|Jan 3, 2011

Vehicle License Plate Recognition13 references13 citations

TL;DR

This paper presents an efficient character segmentation method for business card images captured by mobile phone cameras, using skew correction followed by horizontal and vertical histogram-based line and character segmentation. It achieves 97.48% mean segmentation accuracy on 3-megapixel images within 1.1 seconds on a moderately powered notebook, making it suitable for resource-constrained mobile devices.

ABSTRACT

Due to huge deformation in the camera captured images, variety in nature of the business cards and the computational constraints of the mobile devices, design of an efficient Business Card Reader (BCR) is challenging to the researchers. Extraction of text regions and segmenting them into characters is one of such challenges. In this paper, we have presented an efficient character segmentation technique for business card images captured by a cell-phone camera, designed in our present work towards developing an efficient BCR. At first, text regions are extracted from the card images and then the skewed ones are corrected using a computationally efficient skew correction technique. At last, these skew corrected text regions are segmented into lines and characters based on horizontal and vertical histogram. Experiments show that the present technique is efficient and applicable for mobile devices, and the mean segmentation accuracy of 97.48% is achieved with 3 mega-pixel (500-600 dpi) images. It takes only 1.1 seconds for segmentation including all the preprocessing steps on a moderately powerful notebook (DualCore T2370, 1.73 GHz, 1GB RAM, 1MB L2 Cache).

Motivation & Objective

• To address the challenge of segmenting text regions in highly deformed, camera-captured business card images on mobile devices.
• To develop a computationally efficient method that handles skew, noise, and varying business card layouts.
• To enable accurate character segmentation under the strict computational constraints of mobile platforms.
• To achieve high segmentation accuracy while maintaining real-time performance on low-end hardware.

Proposed method

• Text regions are first extracted from the camera-captured business card images using preprocessing techniques.
• A computationally efficient skew correction method is applied to correct geometric distortions in the text regions.
• Horizontal and vertical projection histograms are used to segment the corrected text regions into lines and individual characters.
• The method leverages intensity variations and spatial distribution patterns to identify line breaks and character boundaries.
• All preprocessing and segmentation steps are optimized for low computational overhead to suit mobile device deployment.
• The approach is evaluated on 3-megapixel images at 500–600 dpi resolution to reflect real-world mobile capture quality.

Experimental results

Research questions

• RQ1 How can text regions be reliably extracted from highly deformed business card images captured by mobile phone cameras?
• RQ2 What is an efficient skew correction technique suitable for mobile devices with limited processing power?
• RQ3 How can lines and characters be accurately segmented using histogram-based methods under real-world image distortions?
• RQ4 What level of segmentation accuracy can be achieved while maintaining real-time performance on low-end mobile hardware?

Key findings

• The proposed method achieves a mean segmentation accuracy of 97.48% on 3-megapixel business card images captured at 500–600 dpi resolution.
• The entire segmentation pipeline, including preprocessing and skew correction, runs in 1.1 seconds on a moderately powerful notebook (DualCore T2370, 1.73 GHz, 1GB RAM).
• The method is computationally efficient and suitable for deployment on mobile devices with limited computational resources.
• Skew correction significantly improves segmentation performance by aligning text regions before line and character extraction.
• Histogram-based segmentation effectively handles variations in font size, spacing, and layout across diverse business card designs.