A pilot study investigating the rapid estimation of intact dry density of chalk using low impact energy hardness measurements taken with an Equotip Leeb hardness tester

Intact dry density (IDD) is a key parameter for characterising the rock mass behaviour of chalk and is often the main classification property reported for engineering design. The role of IDD in coastal chalk settings has been discussed in literature as an important rock control for coastal cliff recession behaviour. This study demonstrates a tool that can reliably estimate the IDD of chalk samples. Originally developed for testing sheet steel materials, the Equotip has been shown to be adaptable for geological applications and the results of this study show that the Equotip can be used to derive a surface hardness value of dry chalk samples that has a positive correlation with IDD. Furthermore, the results show that the device is sensitive to the anisotropy of chalk and that testing undertaken in the diametral orientation give more accurate correlations. The results of this study also show that the Equotip is sensitive to environmental setting and the effects of weathering, giving a separate correlation for samples collected along the Sussex coastline to those obtained inland. A statistical assessment of the results shows that a characteristic hardness within an acceptable margin of error can be achieved from an optimal number of single impacts.

On determining frost susceptibility of soils

Freezing and thawing of soils are common in cold regions. They may even be considered as limit states from the thermo-hydro-mechanical point of view. Thus, to characterize the behavior of freezing and thawing soils, some basic principles should be considered. Design to prevent frost damage should be based on theories that have been shown to apply to field conditions. The laboratory procedures used in the design should imitate the expected freezing behavior, and the tested specimen should be prepared to simulate the soil conditions in the field. The parameters from the test should be applicable to the design model. If these principles are not applied, then the frost susceptibility can be considered as a limit classification property describing the risk of damage in freezing and thawing.

Object recognition and detection with deep learning for autonomous driving applications

Autonomous driving requires reliable and accurate detection and recognition of surrounding objects in real drivable environments. Although different object detection algorithms have been proposed, not all are robust enough to detect and recognize occluded or truncated objects. In this paper, we propose a novel hybrid Local Multiple system (LM-CNN-SVM) based on Convolutional Neural Networks (CNNs) and Support Vector Machines (SVMs) due to their powerful feature extraction capability and robust classification property, respectively. In the proposed system, we divide first the whole image into local regions and employ multiple CNNs to learn local object features. Secondly, we select discriminative features by using Principal Component Analysis. We then import into multiple SVMs applying both empirical and structural risk minimization instead of using a direct CNN to increase the generalization ability of the classifier system. Finally, we fuse SVM outputs. In addition, we use the pre-trained AlexNet and a new CNN architecture. We carry out object recognition and pedestrian detection experiments on the Caltech-101 and Caltech Pedestrian datasets. Comparisons to the best state-of-the-art methods show that the proposed system achieved better results.

Device-Oriented Automatic Semantic Annotation in IoT

Semantic technologies are the keys to address the problem of information interaction between assorted, heterogeneous, and distributed devices in the Internet of Things (IoT). Semantic annotation of IoT devices is the foundation of IoT semantics. However, the large amount of devices has led to the inadequacy of the manual semantic annotation and stressed the urgency into the research of automatic semantic annotation. To overcome these limitations, a device-oriented automatic semantic annotation method is proposed to annotate IoT devices’ information. The processes and corresponding algorithms of the automatic semantic annotation method are presented in detail, including the information extraction, text classification, property information division, semantic label selection, and information integration. Experiments show that our method is effective for the automatic semantic annotation to IoT devices’ information. In addition, compared to a typical rule-based method, the comparison experiment demonstrates that our approach outperforms this baseline method with respect to the precision and F-measure.

Research on Relative Reducts of Rough Set Model in Consistent and Inconsistent Decision Tables

Z. Pawlak’s classical rough set theory has been widely applied in analyzing ordinary information systems and decision tables. In this theory, a relative reduct can be considered as a minimum set of attributes that preserves a certain classification property. This paper investigates three different classification properties, and proposes three distinct definitions accordingly. According to the three classification properties, we can define three distinct definitions respectively. Based on the common structure of the specific definitions of relative reducts and discernibility matrices, general definitions of relative reducts and discernibility matrices are suggested.

Classification of Sketch Strokes and Corner Detection Using Conic Sections and Adaptive Clustering

This paper presents a method for classifying pen strokes in an on-line sketching system. The method, based on linear least squares fitting to a conic section equation, proposes using the conic equation’s natural classification property to help classify sketch strokes and identify lines, elliptic arcs, and corners composed of two lines with an optional fillet. The hyperbola form of the conic equation is used for corner detection. The proposed method has proven to be fast, suitable for real-time classification, and capable of tolerating noisy input, including cusps and spikes. The classification is obtained in o(n) time in a single path, where n is the number of sampled points. In addition, an improved adaptive method for clustering disconnected end-points is proposed. The notion of in-context analysis is discussed, and examples from a working implementation are given.