Detecting irregular-shaped text instances is the main challenge for text detection. Existing approaches can be roughly divided into top-down and bottom-up perspective methods. The former encodes text contours into unified units, which always fails to fit highly curved text contours. The latter represents text instances by a number of local units, where the complicated network and post-processing lead to slow detection speed. In this paper, to detect arbitrary-shaped text instances with high detection accuracy and speed simultaneously, we propose a \textbf{Bi}directional \textbf{P}erspective strategy based \textbf{Net}work (BiP-Net). Specifically, a new text representation strategy is proposed to represent text contours from a top-down perspective, which can fit highly curved text contours effectively. Moreover, a contour connecting (CC) algorithm is proposed to avoid the information loss of text contours by rebuilding interval contours from a bottom-up perspective. The experimental results on MSRA-TD500, CTW1500, and ICDAR2015 datasets demonstrate the superiority of BiP-Net against several state-of-the-art methods.
翻译:检测异常成形的文本实例是探测文本的主要挑战。 现有的方法可以大致分为上到下到下到下到上到下到上到上到下到上到上到上到上到上到上到上到上到上到上到上到上到上到下到上到上到下到上到上到下到上到上到下到上到上到上到下到上到下到下到上到上到上到下到上到上到上到上到上到下到上到上到上到上到上到上到上到上到上到下到上,后者是一些地方单位的文字实例,复杂的网络和后到后处理导致检测速度缓慢。 在本文中,为了探测到高检测准确性和速度,我们建议了一种任意到的文本概要的连接算法,通过从下到下到下到下到的间距等,来避免对文本轮进行信息损失。 具体来说, 以自上到 MSRA- TD500, CTW1500, CT1500 和 ICDAR- NAD 的几- 和 BISD- 数据集的优越性。