Road Sign Detection in Images: A Case Study

Belaroussi, Rachid; Foucher, Philippe; Tarel, Jean-Philippe; Soheilian, Bahman; Charbonnier, Pierre; Paparoditis, Nicolas

doi:10.1109/icpr.2010.1125

Cited by 65 publications

(30 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…La détection de panneaux fondée sur la TCB et la TSB a été comparée à des approches fondées sur les transformées TSR et transformée de Hough dans (Belaroussi et al, 2010a). Il est ressorti de cette étude que l'accumulation bi-variée est plus efficace qu'une transformation mono-variée lorsque la taille des panneaux est suffisante, typiquement supérieure à 10 % de la taille de l'image.…”

Section: Les Méthodes De Détection De Panneauxunclassified

“…-Base Stéréopolis (Belaroussi et al, 2010a) : 847 images de dimensions 480×540 contenant 10 panneaux triangulaires et 21 panneaux circulaires ou rectangulaires (voir figure 13b). C'est une base publique : ❤tt♣✿✴✴✇✇✇✳✐t♦✇♥s✳❢r✴❜❡♥❝❤♠❛r❦✐♥❣.…”

Section: Les Bases De Testunclassified

“…La méthode présentée a été comparée à deux autres algorithmes issus de l'état de l'art dans (Belaroussi et al, 2010a). Nous montrons que les approches cumulatives bivariées sont plus robustes que les approches monovariées (transformées TSR et transformée de Hough).…”

Section: Conclusion Et Perspectivesunclassified

“…The BCT and VBT approaches are compared to a monovariate cumulative method (radial symmetry transform and Hough transform) in (Belaroussi et al, 2010a). We estimate the performances of these detectors on three different databases of respectively 48 images 360 × 270, 89 images 640 × 480 and 847 images 480 × 540.…”

Section: Extended Abstractmentioning

confidence: 99%

“…The VBT is also quite robust to vertex occlusion and missing edges as shown in figure 2. Indeed, set of voters are taken from the sides of a triangle so if a consistent portion of the triangle edges is still visible, its vertices can be found even if they are occluded.The BCT and VBT approaches are compared to a monovariate cumulative method (radial symmetry transform and Hough transform) in (Belaroussi et al, 2010a). We estimate the performances of these detectors on three different databases of respectively 48 images 360 × 270, 89 images 640 × 480 and 847 images 480 × 540.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Détection des panneaux de signalisation routière par accumulation bivariée

Belaroussi¹,

Tarel²

2010

Traitement du signal

Self Cite

View full text Add to dashboard Cite

RÉSUMÉ. Nous présentons une méthode géométrique utilisant l'orientation du gradient pour la détection de la signalisation verticale dans des images fixes, indépendamment de leur position et de leur orientation. La détection est réalisée par une transformation de type accumulateur de Hough bivariée, fondée sur l'utilisation de paires de points avec des contraintes sur leurs gradients. Les panneaux circulaires et polygonaux (non triangulaires) sont détectés par la transformation chinoise bilatérale TCB. Cette transformation est rapide et ne fait pas de distinction entre les cercles et les polygones 4 côtés ou plus. Le cas des panneaux triangulaires est traité par la transformation en sommet et bissectrice TSB, capable de détecter précisemment les bissectrices et les sommets d'un triangle. Les performances de la TCB et de la TSB sont estimées sur plusieurs bases d'images de scènes urbaines.ABSTRACT. We present a real-time approach for circular and polygonal road signs detection in still images, regardless of their pose and orientation. Object detection is performed using a Hough like transform based on accumulation on pairs of points with their gradients. Circular and polygonal (non triangular) traffic signs are detected by the so-called Bilateral Chinese Transform BCT. Triangular traffic signs are processed by the Vertex Bisector Transform VBT. These two approaches are tested on several databases of urban scene.MOTS-CLÉS : transformation en symétrie généralisée, transformation en symétrie pour les coins, transformation en sommet et bissectrice, transformation chinoise, contours orientés, panneaux routiers, temps réel, transformée de Hough. Extended abstractRoad sign identification in images is an important issue, especially for vehicle safety and road management applications. It is usually tackled in three stages : detection, recognition and tracking, and evaluated as a whole. To progress towards better algorithms, we focus in this article on the first stage of the process, namely road sign detection. We focus our work on a feature-based approach to build geometrical models of various kind of shapes : triangle on the one hand, square, diamond, octagon and circle on the other hand.We introduce two cumulative bivariate transformations :-the Bilateral Chinese Transform BCT, tailored for the detection of circular and polygonal road signs with four or more sides.-The Vertex and Bisector Transformation VBT, dedicated to the detection of triangular traffic signs.In the BCT approach (Belaroussi et al., 2009b), object detection is done using a pairwise gradient-based symmetry transform able to detect circles and polygons indistinctly. The BCT seeks the symmetry of shape in the gradient orientation : it decomposes an object into a set of parallel contours with opposite gradients, and models this gradient field symmetry using an accumulation of radial symmetry evidences. Figure 7 illustrates the signed accumulators obtained with the BCT over two elliptical shapes : it is worth noticing that the BCT is also orientation free....

show abstract

Section: Les Méthodes De Détection De Panneauxunclassified

Section: Les Bases De Testunclassified

Section: Conclusion Et Perspectivesunclassified

Section: Extended Abstractmentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Détection des panneaux de signalisation routière par accumulation bivariée

Belaroussi¹,

Tarel²

2010

Traitement du signal

Self Cite

View full text Add to dashboard Cite

show abstract

Uncertainty quantification for deep neural networks: An empirical comparison and usage guidelines

Weiß

Tonella

2023

Software Testing Verif & Rel

View full text Add to dashboard Cite

SummaryDeep neural networks (DNN) are increasingly used as components of larger software systems that need to process complex data, such as images, written texts, audio/video signals. DNN predictions cannot be assumed to be always correct for several reasons, amongst which the huge input space that is dealt with, the ambiguity of some inputs data, as well as the intrinsic properties of learning algorithms, which can provide only statistical warranties. Hence, developers have to cope with some residual error probability. An architectural pattern commonly adopted to manage failure prone components is the supervisor, an additional component that can estimate the reliability of the predictions made by untrusted (e.g., DNN) components and can activate an automated healing procedure when these are likely to fail, ensuring that the deep learning‐based system (DLS) does not cause damages, despite its main functionality being suspended.In this paper, we consider DLS that implement a supervisor by means of uncertainty estimation. After overviewing the main approaches to uncertainty estimation and discussing their pros and cons, we motivate the need for a specific empirical assessment method that can deal with the experimental setting in which supervisors are used, where accuracy of the DNN matters only as long as the supervisor lets the DLS continue to operate. Then we present a large empirical study conducted to compare the alternative approaches to uncertainty estimation. We distilled a set of guidelines for developers that are useful to incorporate a supervisor based on uncertainty monitoring into a DLS.

show abstract

Intensive Survey About Road Traffic Signs Preprocessing, Detection and Recognition

Almuhajri

Suen

2019

Communications in Computer and Information Science

View full text Add to dashboard Cite

Road Sign Detection in Images: A Case Study

Cited by 65 publications

References 13 publications

Détection des panneaux de signalisation routière par accumulation bivariée

Détection des panneaux de signalisation routière par accumulation bivariée

Uncertainty quantification for deep neural networks: An empirical comparison and usage guidelines

Intensive Survey About Road Traffic Signs Preprocessing, Detection and Recognition

Contact Info

Product

Resources

About