Text detection in street level images - Archive ouverte HAL Access content directly
Journal Articles Pattern Analysis and Applications Year : 2013

Text detection in street level images


Text detection system for natural images is a very challenging task in Computer Vision. Image acquisition introduces distortion in terms of perspective, blurring, illumination, and characters may have very diff erent shape, size, and color. We introduce in this article a full text detection scheme. Our architecture is based on a new process to combine a hypothesis generation step to get potential boxes of text and a hypothesis validation step to filter false detections. The hypothesis generation process relies on a new efficient segmentation method based on a morphological operator. Regions are then filtered and classi ed using shape descriptors based on Fourier, Pseudo Zernike moments and an original polar descriptor, which is invariant to rotation. Classi cation process relies on three SVM classi ers combined in a late fusion scheme. Detected characters are finally grouped to generate our text box hypotheses. Validation step is based on a global SVM classi cation of the box content using dedicated descriptors adapted from the HOG approach. Results on the well-known ICDAR database are reported showing that our method is competitive . Evaluation protocol and metrics are deeply discussed and results on a very challenging street-level database are also proposed.
Fichier principal
Vignette du fichier
PAA2013_fabrizio_marcotegui_cord_text_detection.pdf (3.81 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-00906841 , version 1 (20-11-2013)



Jonathan Fabrizio, Beatriz Marcotegui, Matthieu Cord. Text detection in street level images. Pattern Analysis and Applications, 2013, 16 (4), pp.519-533. ⟨10.1007/s10044-013-0329-7⟩. ⟨hal-00906841⟩
4867 View
408 Download



Gmail Facebook X LinkedIn More