Description: 1. We propose a novel text detector for Arbitrary-Shaped Text. The proposed method effectively detects text area by splitting each text region into text center area and text border area.
2. Most scene text detection algorithms are deep learning based methods that depend on various scale of anchors and bounding box regression. But most existing quadrangular bounding box based detectors are difficult to locate texts with arbitrary shapes. So we use semantic segmentation methods to detect arbitrary shapes texts.
3. Semantic segmentation methods may not separate the test instance that are very close to each other. So we splitting each text region into text center area and text border area to make segmentation-based methods to detect arbitrary shapes texts easily.