Axel Brink

Writer identification using directional ink-trace width measurements

As suggested by modern paleography, the width of ink traces is a powerful source of information for off-line writer identification, particularly if combined with its direction. Such measurements can be computed using simple, fast and accurate methods based on pixel contours, the combination of which forms a powerful feature for writer identification: the Quill feature. It is a probability distribution of the relation between the ink direction and the ink width. It was tested in writer identification experiments on two datasets of challenging medieval handwriting and two datasets of modern handwriting. The feature achieved a nearest-neighbor accuracy in the range of 63-95%, which even approaches the performance of two state-of-the-art features in contemporary-writer identification (Hinge and Fraglets). The feature is intuitive and explainable and its principle is supported by a model of trace production by a quill. It illustrates that ink width patterns are valuable. A slightly more complex variant of Quill, QuillHinge, scored 70-97% writer identification accuracy. The features are already being used by domain experts using a graphical interface.

Text-Independent Writer Identification and Verification on Offline Arabic Handwriting

In this paper, we evaluate the performance on Arabic handwriting of the text-independent writer identification methods that we developed and tested on Western script in recent years. We use the IFN/ENIT data in the experiments reported here and our tests involve 350 writers. The results show that our methods are very effective and the conclusions drawn in previous studies remain valid also on Arabic script. High performance is achieved by combining textural features (joint directional probability distributions) with allographic features (grapheme-emission distributions).

Towards Explainable Writer Verification and Identification Using Vantage Writers

In this paper, a new method for off-line writer verification and identification is proposed which encodes writer features as a mix of typical handwriting styles, written by so-called vantage writers. Since their handwriting can be shown to the user, the method provides a degree of transparency that is usually not present in automatic verification and identification systems. It acts as a dimensionality reduction of a precomputed basic feature vector. In this experiment, the hinge feature was used. The method was tested with unconstrained connected cursive text in two datasets: a known dataset of 252 writers and 1074 writers from a new, forensic dataset.

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

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.

Towards robust writer verification by correcting unnatural slant

Slant is a salient feature of Western handwriting and it is considered to be an important writer-specific feature. In disguised handwriting however, slant is often modified. It was tested whether slant is indeed an important factor and it was tested whether the distorting effect of deliberate slant change can be countered by a simple shear transform. This was done in two off-line writer verification experiments in image processing conditions of slant elimination and slant correction. The experiments were performed using three features based on statistical pattern recognition, including the state-of-the-art features Fraglets and Hinge. A new public dataset was created and used, containing natural and slanted handwriting by 47 writers. A striking result is that the average natural slant value is much less important for biometric systems than is usually assumed: eliminating slant yields just a 1-5% performance loss. A second result is that the effects of deliberate slant change cannot be fully countered by a simple shear transform: it raises performance on the distorted handwriting from 53-68% to 64-90%, but this is still lower than normal operation on natural handwriting: 97-100%.

Recognition of Handwritten Numerical Fields in a Large Single-Writer Historical Collection

This paper presents a segmentation-based handwriting recognizer and the performance that it achieves on the numerical fields extracted from a large single-writer historical collection. Our recognizer has the particularity that it uses morphing during training: random elastic deformations are applied to fabricate synthetic training character patterns yielding an improved final recognition performance. Two different digit recognizers are evaluated, a multilayer perceptron (MLP) and radial basis function network (RBF), by plugging them into the same left-to-right Viterbi search framework with a tree organization of there cognition lexicon. We also compare with the performance obtained when no dictionary is used to constrain the recognition results.

Automatic removal of crossed-out handwritten text and the effect on writer verification and identification

A method is presented for automatically identifying and removing crossed-out text in off-line handwriting. It classifies connected components by simply comparing two scalar features with thresholds. The performance is quantified based on manually labeled connected components of 250 pages of a forensic dataset. 47% of connected components consisting of crossed-out text can be removed automatically while 99% of the normal text components are preserved. The influence of automatically removing crossed-out text on writer verification and identification is also quantified. This influence is not significant.

Reducing the Time Complexity and Identifying Ill-Posed Problem Instances of Minkowski Sum Based Similarity Calculations

To calculate the Minkowski-sum based similarity measure of two convex polyhedra, many relative orientations have to be considered. These relative orientations are characterized by the fact that some faces and edges of the polyhedra are parallel. For every relative orientation of the polyhedra, the volume or mixed volume of their Minkowski sum is evaluated. From the minimum of this volume, the similarity measure is calculated. In this article two issues are addressed. First, we propose and test a method to reduce the set of relative orientations to be considered by using geometric inequalities in the slope diagrams of the polyhedra. In this way, the time complexity of O(n6) is reduced to O(n4.5). Secondly, we determine which relative orientation problems are ill-posed and may be skipped because they do not maximize the similarity measure.

Interactive evolutionary computing for the binarization of degenerated handwritten images

The digital cleaning of dirty and old documents and the binarization into a black/white image can be a tedious process. It is usually done by experts. In this article a method is shown that is easy for the end user. Untrained persons are able to do this task now while before an expert was needed. The method uses interactive evolutionary computing to program image processing operations that act on the document image.

Reducing the Time Complexity of Minkowski-Sum Based Similarity Calculations by Using Geometric Inequalities

The similarity of two convex polyhedra A and B may be calculated by evaluating the volume or mixed volume of their Minkowski sum over a specific set of relative orientations. The relative orientations are characterized by the fact that faces and edges of A and B are parallel as much as possible. For one of these relative orientations the similarity measure is optimal. In this article we propose and test a method to reduce the number of relative orientations to be considered by using geometric inequalities in the slope diagrams of A and B. In this way the time complexity of O(n 6) is reduced to O(n 4.5). This is derived, and verified experimentally.