Tekleyohannes Menbere Kina, Rybalkin Vladimir, Ghaffar Muhammad Mohsin, Varela Javier Alejandro, Wehn Norbert, Dengel Andreas
Microelectronic Systems Design Research Group, University of Kaiserslautern, 67663 Kaiserslautern, Germany.
German Research Center for Artificial Intelligence (DFKI), 67663 Kaiserslautern, Germany.
J Imaging. 2021 Sep 3;7(9):175. doi: 10.3390/jimaging7090175.
In recent years, there has been an increasing demand to digitize and electronically access historical records. Optical character recognition (OCR) is typically applied to scanned historical archives to transcribe them from document images into machine-readable texts. Many libraries offer special stationary equipment for scanning historical documents. However, to digitize these records without removing them from where they are archived, portable devices that combine scanning and OCR capabilities are required. An existing end-to-end OCR software called anyOCR achieves high recognition accuracy for historical documents. However, it is unsuitable for portable devices, as it exhibits high computational complexity resulting in long runtime and high power consumption. Therefore, we have designed and implemented a configurable hardware-software programmable SoC called DocChip that makes use of anyOCR techniques to achieve high accuracy. As a low-power and energy-efficient system with real-time capabilities, the DocChip delivers the required portability. In this paper, we present the hybrid CPU-FPGA architecture of DocChip along with the optimized software implementations of the anyOCR. We demonstrate our results on multiple platforms with respect to runtime and power consumption. The DocChip system outperforms the existing anyOCR by 44× while achieving 2201× higher energy efficiency and a 3.8% increase in recognition accuracy.
近年来，对历史记录进行数字化和电子访问的需求日益增长。光学字符识别（OCR）通常应用于扫描的历史档案，以便将它们从文档图像转录为机器可读文本。许多图书馆提供用于扫描历史文档的专用固定设备。然而，为了在不将这些记录从存档位置移除的情况下进行数字化，需要具备扫描和OCR功能的便携式设备。现有的一款名为anyOCR的端到端OCR软件对历史文档具有较高的识别准确率。然而，它不适用于便携式设备，因为它具有较高的计算复杂度，导致运行时间长且功耗高。因此，我们设计并实现了一种名为DocChip的可配置硬件 - 软件可编程片上系统（SoC），它利用anyOCR技术来实现高精度。作为一个具有实时能力的低功耗且节能的系统，DocChip具备所需的便携性。在本文中，我们介绍了DocChip的混合CPU - FPGA架构以及anyOCR的优化软件实现。我们在多个平台上展示了关于运行时间和功耗的结果。DocChip系统的性能比现有的anyOCR高出44倍，同时实现了高2201倍的能源效率，并且识别准确率提高了3.8%。
