Skew Detection/Correction and Local Minima/Maxima Techniques for Extracting a New Arabic Benchmark Database
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2015, Vol 6, Issue 3
Abstract
We propose a set of techniques for extracting a new standard benchmark database for Arabic handwritten scripts. Thresholding, filtering, and skew detection/correction techniques are developed as a pre-processing step of the database. Local minima and maxima using horizontal and vertical histogram are implemented for extracting the script elements of the database. Elements of the database contain pages, paragraphs, lines, and characters. The database divides into two major parts. The first part represents the original elements without modifications; the second part represents the elements after applying the proposed techniques. The final database has collected, extracted, validated, and saved. All techniques are tested for extracting and validating the elements. In this respect, ACDAR proposes a first issue of the Arabic benchmark databases. In addition, the paper confirms establishment a specialized research-oriented center refers to learning, teaching, and collaboration activities. This center is called "Arabic Center for Document Analysis and Recognition (ACDAR)" which is similar to other centers developed for other languages such as English.
Authors and Affiliations
Husam Al Hamad
Keywords
Related Articles
- EP ID EP153143
- DOI 10.14569/IJACSA.2015.060301
- Views 126
- Downloads 0
A New Task Scheduling Algorithm using Firefly and Simulated Annealing Algorithms in Cloud Computing
Task scheduling is a challenging and important issue, which considering increases in data sizes and large volumes of data, has turned into an NP-hard problem. This has attracted the attention of many researchers througho...
Level of Confidence in Software Effort Estimation by an Intelligent Fuzzy – Neuro - Genetic Approach
Organizations are struggling to deliver the expected software functionality and quality in scheduled time and prescribed budget. Despite availability of numerous advanced effort estimation techniques overestimation and u...
On Some Methods for Dimensionality Reduction of ECG Signals
Dimensionality reduction with two methods, namely, Laplacian Eigenmap (LE) and Locality Preserving Projections (LPP) is studied for normal and pathological noisy and noiseless ECG patterns. Besides, the possibility of us...
A Tool Design of Cobit Roadmap Implementation
Over the last two decades, the role of information technology in organizations has changed from primarily a supportive and transactional function to being an essential prerequisite for strategic value generation. The org...
Software Abstractions for Large-Scale Deep Learning Models in Big Data Analytics
The goal of big data analytics is to analyze datasets with a higher amount of volume, velocity, and variety for large-scale business intelligence problems. These workloads are normally processed with the distribution on...