Scene Text Segmentation in the KAIST Dataset

Datasets -> Datasets List -> Current Page

Created: 2010-01-25 Last updated: 2011-001-27

Description

The goal of this task is to automatically separate text pixels from background pixels in the KAIST dataset images. In other words, every pixel has to be classified as text pixel or background pixel in this task. Localisation information should not be assumed known.

The evaluation method to be used should be the calculation of precision and recall metrics at the pixel level.

Related Dataset

• KAIST Scene Text Database

Related Ground Truth Data

• KAIST Scene Text Ground Truth (text location, segmantation and recognition)

Related Software

None

References

---

This page is editable only by TC11 Officers .