A Robust Software Barcode Reader Using the Hough Transform . In this paper we present a method based on the Hough transform which. Published in: · Proceeding. ICIIS ’99 Proceedings of the International Conference on Information Intelligence and Systems. Page March 31 – April A Robust Software Barcode Reader Using the Hough Transform (Englisch). Muniz, R. / Junco, L. / Otero, A. / Institute of Electrical and Electronics Engineers.

Author: Garisar Mosida
Country: South Africa
Language: English (Spanish)
Genre: Science
Published (Last): 24 June 2008
Pages: 216
PDF File Size: 13.24 Mb
ePub File Size: 20.42 Mb
ISBN: 357-4-55859-146-2
Downloads: 54430
Price: Free* [*Free Regsitration Required]
Uploader: Doktilar

These points are extracted by thresholding the transformed image taking a suitable threshold.

An example of deformed model is shown in Fig. The consistency checking with the constraint network rectifies the effect of the noise by checking the relationships of the widths of the nearby bars and taking into consideration of the decoding rules of EAN Algorithm outline First, based on the previously detected endpoints of the scanline, we compute the spatial location of each digit segment in the barcode.

Unfortunately, we could find only one barcode image database for comparative assessment 1. Please review our privacy policy. Unlike previous approaches, our algorithm does not binarize the image, thus sidestepping a critical and often error-prone, early-commitment procedure. Chai and Hock use a single scanline binarized using the average gray level as a threshold [ 4 ] whereas Wachenfeld et al. As a result, our approach is capable of decoding 1-D barcodes from noisy pictures even in the presence of motion blur or lack focus.

Barcode reading needs to be robust to challenging conditions such as blur, noise, low resolution, or low quality camera lenses, all of which are extremely common.

In contrast, our decoding algorithm never performs a binarization or an edge extraction operation. If present it will move on to uusing full search and sotware to decode trasform. In our implementation we assume a localization tolerance in either end point equal to twice the width of the narrowest bar.


A system that binarizes the intensity would be hard-pressed to detect the correct pattern. Processor Utilization for the Real-Time Implementation. Although deformable template matching usually requires costly optimization techniques, we prove that in the case of these patterns, matching can be obtained exactly with a simple discrete search.

Commercial scanners, such as those used in supermarkets, shine a stripe of pulsed light on the code and measure the intensity of its reflection; the use of active illumination makes them virtually insensitive to changes of ambient illumination.

Reading 1-D Barcodes with Mobile Phones Using Deformable Templates

Uxing variety of functions can be considered for the log-likelihood D modeling the discrepancy between model and observation. A UPC-A barcode encodes 12 symbols each representing a number 0 — 9. Our derivation exploits the fact that D IM ow k is piecewise constant in the ow space.

In this case, the integral is proportional to the area of the polygonal cell, which can be easily computed and stored offline.

We define a global cost function that, for each possible sequence of symbols, penalizes overlaps or gaps in the sequence of deformed templates, with the deformation parameters obtained by least squares regression.

Localization is only a function of the image size and it takes approximately ms for a VGA image. One important contribution of this work is to derive an algorithm to compute this integral exactly and in affordable computing time.

We employ a particular form ksing deformable template matching, that produces robust results even in difficult situations, such as the one shown in Fig. This method showed an improvement of over a factor of three, of the frame rate. Digit segment — total likelihood In order to compute the likelihood of an observed scanline for given symbol, it houh necessary to take the uncertainty about o and w into consideration. This is achieved by scanning through multiple lines of the barcode image.


We now describe in detail each of the steps of this procedure. In order to assess our system in other realistic situations, we gathered a number of images taken from two different cellphones, tramsform created three new data sets. A sample of images correctly decoded by our algorithm is shown in Fig. A model is therefore an archetypal representation of one symbol of a standardized robjst, plus one bar from each one of the nearby symbols. Until recently, however, barcode reading was performed almost exclusively with dedicated hardware.

Therefore, in order to localize the endpoints o L and o R of the barcode, we first determine the intersections i L and i R of the scanline l n with w rectangle and then, robbust rectangle being larger than the actual barcode, we proceed inwards from each end see Fig. A challenging barcode image that is correctly decoded by our algorithm. We also made the data sets public to allow other researchers to test and compare their methods.

The maximum rotation angle for correct localization is bound by simple geometric considerations.

Reading 1-D Barcodes with Mobile Phones Using Deformable Templates

Due to the presence of the noise, there are some infrequent narrow peaks and valleys appeared in the waveform. Original Image Figure 9: Supplementary Material Dataset used for the experiments Click here to view.

In particular, the image may be of poor quality, due to noise and possibly low contrast. The minimum cost sequence can then be found via dynamic programming. A parameterized model is a shifted and scaled deformed version of th original model:.

Our technique relies on deformable templates and exploits all the gray level information of each pixel.