In this paper, we present a pattern recognition method that uses dynamic programming for the alignment of Radon features. The key characteristic is to use time warping (DTW) match corresponding pairs features all possible projections. Thanks DTW, avoid compressing feature matrix into single vector which would otherwise miss information. To reduce number matchings, rely on initial normalization based orientation. A comprehensive study made using major state-of-the-art shape descriptors over...

In this paper, we present a fully operational, scalable and open architecture allowing end-to-end document analysis benchmarking without needing to develop the whole pipeline. By decomposing process into coarse-grained tasks, by building upon community provided state-of-the art algorithms, our allows any combination of elementary regardless their running system environment, programming language or data structures. Its flexible structure makes it straightforward plug in new compare them other...

In this paper, we address a new scheme for symbol retrieval based on bag-of-relations (BoRs) which are computed between extracted visual primitives (e.g. circle and corner). Our features consist of pairwise spatial relations from all possible combinations individual primitives. The key characteristic the overall process is to use topological relation information indexed in BoRs recognition. As consequence, directional matching takes place only with those candidates having similar...

In this paper, we propose a new scheme for Devanagari natural handwritten character recognition. It is primarily based on spatial similarity-based stroke clustering. A feature of consists string pen-tip positions and directions at every position along the trajectory. uses dynamic time warping algorithm to align strokes with stored templates determine their similarity. Experiments are carried out help 25 native writers recognition rate approximately 95% achieved. Our recognizer robust large...

In this paper, we discuss how the focus in document analysis graphics recognition has moved from re-engineering problems to indexing and information retrieval. After a few reviews of ongoing work on these topics, propose some challenges for years come.

In this paper, we make an attempt to use inductive logic programming (ILP) automatically learn non trivial descriptions of symbols, based on a formal description. This work is first step in direction and rather proof concept, than fully operational robust framework. The overall goal our approach express graphic symbols by number primitives that may be any complexity (i.e. not necessarily just lines or points) connecting relationships can deduced from straightforward state-of-the art image...

We address visual servoing through a fixed stereo rig using an image Jacobian. Existing methods are based on the stacking of monocular servo Jacobians, resulting in largely over-constrained control commands. In this paper we formally show that epipolar constraint between two images can be taken into account explicitly. then use significantly increases quality task execution, especially where precision, robustness and smoothness movement is concerned.

This paper presents a syntactic recognition approach for on-line drawn graphical symbols. The proposed method consists in an incremental predictive parser based on symbol descriptions by adjacency grammar. analyzes input strokes as they are the user and is able to get ahead which symbols likely be recognized when partial subshape intermediate state. In addition, takes into account two issues. First, any order second, since it framework, system requires real-time response. has been applied...

In this paper we present a method of robustly detect cir- cles in line drawing image. The is fast, robust and very reliable, capable assessing the quality its detection. It based on Random Sample Consensus mini- mization, uses techniques that are inspired from object tracking image sequences. Note : some details illustrations colour. Reading them greylevel printout will reduce their intelligibility.

Making datasets available for peer reviewing of published document analysis methods or distributing large commonly used corpora benchmarking are extremely useful and sound practices initiatives. This paper shows that they cover only a very tiny segment the uses shared research data may have. We develop completely new paradigm sharing accessing common sets, benchmarks other tools is based on open free community contribution model. The model operational has been implemented so it can be tested...

In this paper we present an innovative approach to automatically generate adjacency grammars describing graphical symbols. A grammar production is formulated in terms of rulesets geometrical constraints among symbol primitives. Given a set instances sketched by user using digital pen, our infers the productions consisting ruleset most likely occur. The performance work evaluated comprehensive benchmarking database on-line

In this paper, we present an innovative approach to integrate spatial relations in stroke clustering for handwritten Devanagari character recognition. It handles strokes of any number and order, writer independently. Learnt are hierarchically agglomerated via Dynamic Time Warping based on their location stored accordingly. We experimentally validate our concept by showing its ability improve recognition performance previously published results.

In this paper, we address the use of unified spatial relations for symbol description. We present a topologically guided directional relation signature. It references unique point set instead one entity in pair, thus avoiding problems related to erroneous choices reference entities and preserves symmetry. experimentally validate our method on showing its ability serve retrieval application, based only relational descriptor that represents links between decomposed structural patterns called...