Detecting text in natural scenes with stroke width transform

Epshtein, Boris; Ofek, Eyal; Wexler, Yonatan

doi:10.1109/cvpr.2010.5540041

Cited by 1,239 publications

(843 citation statements)

References 17 publications

Supporting

Mentioning

826

Contrasting

Unclassified

Order By: Relevance

“…The most challenging sub-problem in this system is the text detection and character recognition itself (a problem that has warranted much prior work [6], [7], [8], [9], [10]). Unfortunately, while some off-the-shelf components are available, their performance is generally low when applied to nondocument images (such as those acquired from our robot).…”

Section: Introductionmentioning

confidence: 99%

Autonomous sign reading for semantic mapping

Case

Suresh

Coates

et al. 2011

2011 IEEE International Conference on Robotics and Automation

View full text Add to dashboard Cite

Abstract-We consider the problem of automatically collecting semantic labels during robotic mapping by extending the mapping system to include text detection and recognition modules. In particular, we describe a system by which a SLAMgenerated map of an office environment can be annotated with text labels such as room numbers and the names of office occupants. These labels are acquired automatically from signs posted on walls throughout a building. Deploying such a system using current text recognition systems, however, is difficult since even state-of-the-art systems have difficulty reading text from non-document images. Despite these difficulties we present a series of additions to the typical mapping pipeline that nevertheless allow us to create highly usable results. In fact, we show how our text detection and recognition system, combined with several other ingredients, allows us to generate an annotated map that enables our robot to recognize named locations specified by a user in 84% of cases.

show abstract

Section: Introductionmentioning

confidence: 99%

Autonomous sign reading for semantic mapping

Case

Suresh

Coates

et al. 2011

2011 IEEE International Conference on Robotics and Automation

View full text Add to dashboard Cite

show abstract

“…Lastly, our algorithm output more than just the skeleton, since it also recover the thickness of the stroke in each point, this is why we will refer to the skeleton detected by our algorithm as stroke. We follow the same assumption than in [1], that a stroke used to draw a character has a relatively constant thickness, meaning that the two opposite contours of a stroke are parallels and that the perpendicular of the tangent of a point of one of the contour is also the perpendicular of the tangent on the intersection point of the other contour. Obviously, this assumption is theoretical and in practice the rule is relaxed.…”

Section: Parallel Contour Matching For Strokes Detectionmentioning

confidence: 99%

“…In order for a computer to automatically read text from natural images, two sets of algorithms are required: the first step is to locate where in the images the text is located [1,2], while the second step is about recognising the characters. Our focus in this paper is on the second aspect of the problem.…”

Section: Detection and Recognition Of Text In Natural Imagesmentioning

confidence: 99%

“…Since the algorithm presented in this paper has many similarity with strokewidth transform algorithm [1], it should also be possible to use this stroke detection algorithm to detect where the text is located in natural images. It would be possible to segment the image, then to transform each resulting segment mask into its stroke representation, filter out characters that have a too high variation in thickness and try to recognise the remaining characters.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Strokes Detection for Skeletonisation of Characters Shapes

Berger

2014

Advances in Visual Computing

View full text Add to dashboard Cite

Abstract. Skeletonisation is a key process in character recognition in natural images. Under the assumption that a character is made of a stroke of uniform colour, with small variation in thickness, the process of recognising characters can be decomposed in the three steps. First the image is segmented, then each segment is transformed into a set of connected strokes (skeletonisation), which are then abstracted in a descriptor that can be used to recognise the character. The main issue with skeletonisation is the sensitivity with noise, and especially, the presence of holes in the masks. In this article, a new method for the extraction of strokes is presented, which address the problem of holes in the mask and does not use any parameters.

show abstract

“…The stroke width transform (SWT) [5] and maximally stable extremal regions (MSERs) [6] [16] are two common connected componentbased methods. MSER-based detectors exploit the fact that characters normally have strong contrast with the background to allow for easy reading [3] [19].…”

Section: Related Workmentioning

confidence: 99%

A Text Recognition and Retrieval System for e-Business Image Management

Zhou

McGuinness

O’Connor

2018

MultiMedia Modeling

View full text Add to dashboard Cite

Abstract. The on-going growth of e-business has resulted in companies having to manage an ever increasing number of product, packaging and promotional images. Systems for indexing and retrieving such images are required in order to ensure image libraries can be managed and fully exploited as valuable business resources. In this paper, we explore the power of text recognition for e-business image management and propose an innovative system based on photo OCR. Photo OCR has been actively studied for scene text recognition but has not been exploited for e-business digital image management. Besides the well known difficulties in scene text recognition such as various size, location, orientation in text and cluttered background, e-business images typically feature text with extremely diverse fonts, and the characters are often artistically modified in shape, colour and arrangement. To address these challenges, our system takes advantage of the combinatorial power of deep neural networks and MSER processing. The cosine distance and n-gram vectors are used during retrieval for matching detected text to queries to provide tolerance to the inevitable transcription errors in text recognition. To evaluate our proposed system, we prepared a novel dataset designed specifically to reflect the challenges associated with text in e-business images. We compared our system with two other approaches for scene text recognition, and the results show our system outperforms other state-ofthe-art on the new challenging dataset. Our system demonstrates that recognizing text embedded in images can be hugely beneficial for digital asset management.

show abstract

Detecting text in natural scenes with stroke width transform

Cited by 1,239 publications

References 17 publications

Autonomous sign reading for semantic mapping

Autonomous sign reading for semantic mapping

Strokes Detection for Skeletonisation of Characters Shapes

A Text Recognition and Retrieval System for e-Business Image Management

Contact Info

Product

Resources

About