Robust table recognition for printed document images.

The recognition and analysis of tables on printed document images is a popular research field of the pattern recognition and image processing. Existing table recognition methods usually require high degree of regularity, and the robustness still needs significant improvement. This paper focuses on a robust table recognition system that mainly consists of three parts: Image preprocessing, cell location based on contour mutual exclusion, and recognition of printed Chinese characters based on deep learning network. A table recognition app has been developed based on these proposed algorithms, which can transform the captured images to editable text in real time. The effectiveness of the table recognition app has been verified by testing a dataset of 105 images. The corresponding test results show that it could well identify high-quality tables, and the recognition rate of low-quality tables with distortion and blur reaches 81%, which is considerably higher than those of the existing methods. The work in this paper could give insights into the application of the table recognition and analysis algorithms.

Angle aided circle detection based on randomized Hough transform and its application in welding spots detection.

The Hough transform has been widely used in image analysis and digital image processing due to its capability of transforming image space detection to parameter space accumulation. In this paper, we propose a novel Angle-Aided Circle Detection (AACD) algorithm based on the randomized Hough transform to reduce the computational complexity of the traditional Randomized Hough transform. The algorithm ameliorates the sampling method of random sampling points to reduce the invalid accumulation by using region proposals method, and thus significantly reduces the amount of computation. Compared with the traditional Hough transform, the proposed algorithm is robust and suitable for multiple circles detection under complex conditions with strong anti-interference capacity. Moreover, the algorithm has been successfully applied to the welding spot detection on automobile body, and the experimental results verifies the validity and accuracy of the algorithm.

Modular design and development methodology for robotic multi-axis F/M sensors.

Accurate Force/Moment (F/M) measurements are required in many applications, and multi-axis F/M sensors have been utilized a wide variety of robotic systems since 1970s. A multi-axis F/M sensor is capable of measuring multiple components of force terms along x-, y-, z-axis (Fx, Fy, Fz), and the moments terms about x-, y- and z-axis (Mx, My and Mz) simultaneously. In this manuscript, we describe experimental and theoretical approaches for using modular Elastic Elements (EE) to efficiently achieve multi-axis, high-performance F/M sensors. Specifically, the proposed approach employs combinations of simple modular elements (e.g. lamella and diaphragm) in monolithic constructions to develop various multi-axis F/M sensors. Models of multi-axis F/M sensors are established, and the experimental results indicate that the new approach could be widely used for development of multi-axis F/M sensors for many other different applications.