Search

Document scanning often suffers from skewing, which may seriously influence the efficiency of Optical Character Recognition (OCR). Therefore, it is necessary to correct the skewed document before document image information analysis. In this article, we propose a novel adaptive deskewing algorithm for document images, which mainly includes Skeleton Line Detection (SKLD), Piecewise Projection Profile (PPP), Morphological Clustering (MC), and the image classification method. The image type is determined firstly based on the image's layout feature. Thus, adaptive correcting is applied to deskew the image according to its type. Our method maintains high accuracy on the Document Image Skew Estimation Contest (DISEC'2013) and PubLayNet datasets, which achieved 97.6% and 80.1% accuracy, respectively. Meanwhile, extensive experiments show the superiority of the proposed algorithm.

This paper proposes the Lock-Position-Based RFID Adaptive Parallel Collision Tree (LAPCT) algorithm to address the issues of excessive time slots required in the identification process of collision tree algorithms for multiple tags and the high communication complexity between the reader and multiple tags. The LAPCT algorithm adopts a single-query multiple-response mechanism and dynamically divides the response sub-cycle numbers in the identification cycle based on an adaptive strategy. It uses Manchester encoding to lock collision positions and generate a common query prefix, effectively reducing the number of reader queries. This reduction in queries decreases the total number of required time slots and transmitted bits during the reader–tag communication process, thereby improving the efficiency of multiple tag recognition. Theoretical and simulation experiments demonstrate that compared to similar algorithms, the LAPCT algorithm achieves a maximum reduction of 37% in total time slots required, a maximum improvement of 30% in recognition efficiency, and a maximum reduction of 90% in communication complexity. Furthermore, with an increase in the number of tags, the performance advantages of the LAPCT algorithm become more pronounced, making it suitable for large-scale tag scenarios. [ABSTRACT FROM AUTHOR]

This work focuses on the implementation of innovative adaptive strategies and a closed-loop chain in a piston-damage-based combustion controller. In the previous paper (Part 1), implemented models and the open loop algorithm are described and validated by reproducing some vehicle maneuvers at the engine test cell. Such controller is further improved by implementing self-learning algorithms based on the analytical formulations of knock and the combustion model, to update the fuel Research Octane Number (RON) and the relationship between the combustion phase and the spark timing in real-time. These strategies are based on the availability of an on-board indicating system for the estimation of both the knock intensity and the combustion phase index. The equations used to develop the adaptive strategies are described in detail. A closed-loop chain is then added, and the complete controller is finally implemented in a Rapid Control Prototyping (RCP) device. The controller is validated with specific tests defined to verify the robustness and the accuracy of the adaptive strategies. Results of the online validation process are presented in the last part of the paper and the accuracy of the complete controller is finally demonstrated. Indeed, error between the cyclic and the target combustion phase index is within the range ±0.5 Crank Angle degrees (°CA), while the error between the measured and the calculated maximum in-cylinder pressure is included in the range ±5 bar, even when fuel RON or spark advance map is changing. [ABSTRACT FROM AUTHOR]