Text line detection based on cost optimized local text line direction estimation

Guo, Yandong; Sun, Yufang; Bauer, Péter; Allebach, Jan P.; Bouman, Charles A.

doi:10.1117/12.2083709

Cited by 6 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different from some part-based methods, e.g. [23], TextBoxes++ is hard to fit the accurate boundary for curved texts due to the limitation of quadrilateral representation. Note that all these difficulties also hold for the other state-of-the-art methods [39], [40].…”

Section: F Weaknessesmentioning

confidence: 99%

“…Other popular examples of this type are the works in [19]- [23]; 2) Word-based: Words are directly extracted in the similar manner as general object detection [8], [24]- [26]. In the representative work [8], the authors propose an R-CNN-based [14] framework, where word candidates are first generated with class-agnostic proposal generators followed by a random forest classifier.…”

Section: B Text Detectionmentioning

confidence: 99%

See 1 more Smart Citation

TextBoxes++: A Single-Shot Oriented Scene Text Detector

Liao

Shi

Bai

2018

IEEE Trans. on Image Process.

783

366

View full text Add to dashboard Cite

Scene text detection is an important step of scene text recognition system and also a challenging problem. Different from general object detection, the main challenges of scene text detection lie on arbitrary orientations, small sizes, and significantly variant aspect ratios of text in natural images. In this paper, we present an end-to-end trainable fast scene text detector, named TextBoxes++, which detects arbitrary-oriented scene text with both high accuracy and efficiency in a single network forward pass. No post-processing other than an efficient non-maximum suppression is involved. We have evaluated the proposed TextBoxes++ on four public datasets. In all experiments, TextBoxes++ outperforms competing methods in terms of text localization accuracy and runtime. More specifically, TextBoxes++ achieves an f-measure of 0.817 at 11.6fps for 1024×1024 ICDAR 2015 Incidental text images, and an f-measure of 0.5591 at 19.8fps for 768×768 COCO-Text images. Furthermore, combined with a text recognizer, TextBoxes++ significantly outperforms the stateof-the-art approaches for word spotting and end-to-end text recognition tasks on popular benchmarks. Code is available at: https://github.com/MhLiao/TextBoxes plusplus.

show abstract

Section: F Weaknessesmentioning

confidence: 99%

Section: B Text Detectionmentioning

confidence: 99%

TextBoxes++: A Single-Shot Oriented Scene Text Detector

Liao

Shi

Bai

2018

IEEE Trans. on Image Process.

783

366

View full text Add to dashboard Cite

show abstract

“…Finally, the hybrid methods try to incorporate the local and global information to form text lines. For example, the method in [8] extracts the CCs as symbols. Based on this local information, the direction of the text line is estimated using a cost optimisation framework.…”

Section: Introductionmentioning

confidence: 99%