Sparse-parametric writer identification using heterogeneous feature groups

Schomaker, Lambert; Bulacu, M.; Erp, Merijn van

doi:10.1109/icip.2003.1247019

Cited by 16 publications

(19 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [4], the authors have done some work on edge based directional probabilities distributions as features. In continuation to their previous work, the same authors [17], [15], [16] have extended their work to include multiple heterogeneous sparse groups and edge fragments for writer recognition. Recently, in [3], they have extended their work in this domain and have proposed to use multiple text independent features for writer recognition.…”

Section: Introductionmentioning

confidence: 90%

Writer Identification Using Steered Hermite Features and SVM

Imdad

Bres

Églin

et al. 2007

Ninth International Conference on Document Analysis and Recognition (ICDAR 2007) Vol 2

View full text Add to dashboard Cite

Writer recognition is considered as a difficult problem to solve due to variations found in the writing, even from the same writer. In this paper, Steered Hermite Features are used to identify writer from a written document. We will show that Steered Hermite Features are highly useful for text images because they extract lot of information, notably for data characterized by oriented features, curves and segments. The algorithm we propose here, first calculates the Steered Hermite Features of the images which are then passed on to Support Vector Machine for training and testing. The base of tests consists of sample of some lines of writings (five at most) of primarily diversified writings of authors from IAM database. With the proposed algorithm based on Steered Hermite Features, we were able to achieve an accuracy of around 83% percent for a set of 30 authors with non overlapping images of written text.

show abstract

Section: Introductionmentioning

confidence: 90%

Writer Identification Using Steered Hermite Features and SVM

Imdad

Bres

Églin

et al. 2007

Ninth International Conference on Document Analysis and Recognition (ICDAR 2007) Vol 2

View full text Add to dashboard Cite

show abstract

“…This assumption is not true: the EDD is actually different for different fonts. This font dependence of the EDD is in fact what we, very successfully, exploited in solving the problem of identifying people based on their handwriting Schomaker et al, 2003) (an interesting biometrics method enjoying renewed interest for its forensic applicability). Figure 11.…”

Section: The Neural Networkmentioning

confidence: 96%

Text Detection and Pose Estimation for a Reading Robot

Bulacu¹,

Ezaki²,

Schomaker³

2008

Motion Planning

View full text Add to dashboard Cite

“…It uses a codebook that was trained on all four pages of the first 100 subjects of the Firemaker dataset. The Brush feature [9] is a histogram of ink intensities at stroke endings. It encodes pen landing and lifting habits.…”

Section: Feature Extractionmentioning

confidence: 99%

“…It encodes pen landing and lifting habits. The HrunW feature [1,9] is a histogram of horizontal white run-lengths. This encodes within-and between-letter spacing.…”

Section: Feature Extractionmentioning

confidence: 99%

How much handwritten text is needed for text-independent writer verification and identification

Brink

Bulacu

Schomaker

2008

2008 19th International Conference on Pattern Recognition

View full text Add to dashboard Cite

The performance of off-line text-independent writer verification and identification increases when the documents contain more text. This relation was examined by repeatedly conducting writer verification and identification performance tests while gradually increasing the amount of text on the pages. The experiment was performed on the datasets Firemaker and IAM using four different features. It was also determined what the influence of an unequal amount of text in the documents is. For the best features, it appears that the minimum amount of needed text is about 100 characters.

show abstract

Sparse-parametric writer identification using heterogeneous feature groups

Cited by 16 publications

References 8 publications

Writer Identification Using Steered Hermite Features and SVM

Writer Identification Using Steered Hermite Features and SVM

Text Detection and Pose Estimation for a Reading Robot

How much handwritten text is needed for text-independent writer verification and identification

Contact Info

Product

Resources

About