Sie sind auf Seite 1von 1

Stroke Segmentation for Livestock Brand Image Recognition

Waldemar Villamayor-Venialbo and Horacio Legal-Ayala Laboratorio de Computacin Cient o ca y Aplicada Universidad Nacional de Asuncin, Campus Universitario, San Lorenzo, Paraguay o
wvenialbo@pol.una.py, hlegal@pol.una.py

Abstract In the past year we developed a Cattle Brand Registration System which comprises several Digital Image Processing technologies. Some of the methods that fulll our needs were adopted and incorporated in the system, balancing the trade-o between speed and accuracy. Nevertheless, we faced various problems for which there were not suitable techniques; for instance, no segmentation technique that lls our accuracy requirements was found. This work presents a stroke segmentation approach suitable to perform similarity measuring, using digital image processing techniques, in a cattle brand registration system. The skeletons of the brands are analyzed to detect and separate primary strokes at junction and intersection points. Primary strokes are then mapped into a three-dimensional orientation space to group them into primitive continuous strokes. The results of a set of experiences are presented.

Problem Scenario and Contextualization

Figure 1: A scheme of our Cattle Brand Registration System.

The Cattle Brand Registration System requires new brands to be recorded as sketches handwritten on a digital table. With this digitalization mechanism the strokes in the brand have these characteristics: the width is uniform and not a predominant line feature, the only informative features describing a curve are its orientation and curvature.

Classication of Points, Regions and Strokes


Points and Regions Primary Strokes Primitive Strokes

Figure 2: Several kind of points.

According with the degree of freedom n(P ) of a point P in a brand B: a terminal point T , n(T ) = 1; a normal point N , n(N ) = 2; a T-like junction point J, n(J) = 3; a X-like intersection point I, n(I) = 4; and an isolated dot Z, n(Z) = 0. Points are grouped into: Regular regions R = {P | P B, 0 < n(P ) 2}. Singular regions S = {P | P B, n(P ) 3}.

Figure 3: Several types of primary strokes.

Figure 4: Primitive strokes.

An isolated dot, Z; an isolated stroke, IJK; untied primary strokes, AB, CD, EB, BF , GC, and CH; and a tied primary stroke, BC.

A primitive stroke: the union of one or more primary strokes determining a continuous 2-dimensional curve or a simple polygonal. E.g., ABCD, IJK.

The Method
Stage 1 Primary Stroke Separation 1. Junction and intersection detection
A successful extraction of a primary stroke relies, primarily, on the correct identication of singular points.

Stage 2 Primitive Stroke Reconstruction

Experimental Results
Table 1: Segmentation error rates for two stroke segmentation approaches. One that performs junction analysis on thick lines using the Point to Boundary Orientation Distance (PBOD) technique the detect junction and intersection regions and our approach using the skeleton of the brand.

2. Primary stroke extraction


To extract a primary stroke we proceed isolating each regular region together with each delimiting singular region.

Stage 1 Real Example


Figure 7: Orientation space projection technique.

Stage 2 Real Example A. B.

Junction analysis using Transformation PBOD Skeleton Displacement 13.3 % 8.3 % Rotation 18.1 % 11.4 % Scaling 13.7 % 9.0 % Stretching 14.9 % 9.5 % Shearing 16.2 % 10.7 % Mean Values 15.2 % 9.8 % Bibliography
Chen, J., Sato, Y., Tamura, S.: Orientation space ltering for multiple orientation line segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence 22(5) (2000) 311317 Cao, R., Tan, C.L.: Line primitive extraction by interpretation of line continuation. In: Proceedings of the 6th International Conference on Control, Automation, Robotics and Vision. (2000) Villamayor-Venialbo, W., Legal-Ayala, H.: Stroke segmentation from livestock brand images. In: Proceedings of the 35th Latin American Conference on Informatics. (2009) Villamayor-Venialbo, W., Legal-Ayala, H., Justino, E., Facon, J.: Partial matching using set exclusion criteria: Applied to livestock brand retrieval. In: Proceedings of the 23rd Brazilian Conference on Graphics, Patterns and Images. (2010) 178185

Figure 5: Skeleton of an actual cattle brand (A). Junction and intersection points (B).

Figure 6: Labeled primary strokes and two singular regions, in gray.

Figure 8: Reconnecting primitive strokes.

LCCA Facultad Politcnica Universidad Nacional de Asuncin, Campus Universitario, San Lorenzo, Paraguay e o

http://www.fpuna.edu.py/

Das könnte Ihnen auch gefallen