Applied Mechanics and Materials Vols. 687-691

Abstract: This paper through indoor model test get the axial speed curve and waveform data when the pile top under transient excitation force. Use the method of reverberation ray matrix program to verify the correctness of the velocity curve. The data input to the BP neural network and identification data. It showed that the BP neural network can judge the defect type of pile foundation accurately.

Abstract: In order to implement effective detection and utilize large numbers of unlabeled samples,a pedestrian detection method based on Unsupervised learning was presented.We apply deep learning to human detection to acquire pedestrian features with unlabeled data set.The detection method uses unsupervised convolution sparse auto-encoders to train features at all levels from the data set,then trains classifier with end-to-end supervised method.Additionally,we fine-tune the features in a supervised way.Experiments show that the method approach an state-of-art result on all data set.

Abstract: Aiming at improving the 3D shape measurement speed and accuracy, the design of floating point algorithm for 3-D shape measurement using FPGA, is proposed. By using four-step phase shifting method for measuring 3D measurement system, using CORDIC algorithm to achieve phase calculation of the phase shift method, the algorithm is realized by Verilog language in FPGA, and to make simulation and verification in Modelsim. The study of 3D shape measurement system have a breakthrough from the design performance and hardware.Finally, the MATLAB simulation proved its effectiveness and viability.The experimental results show that the advantages of high data processing speed, high efficiency and high precision,etc.

Abstract: Based on multi-thread technology, completed a set of real-time dynamic measurement system apply on binocular vision principle. Real time measurement require of system efficiency higher, multithreaded programming can greatly improve the efficiency of program. Binocular dynamic real-time detection is based on closed-loop control system, puts forward a multithreaded programming technology which is based on the binocular vision high-speed dynamic measurement method. The data acquisition, processing, display are using multithreading technology.

Abstract: Monitoring the goods’ statue and certificating the identity of the recipient is becoming a big issue in logistics system. In this paper, we propose a system which contains an embedded terminal and a cloud services platform to monitor the real-time statue of the goods. The embedded terminal uses the GPS satellite signal and the nearby base stations to determine the location of the goods. Besides, acceleration sensor and the video capture unit are adopted to monitor the integrity of the goods. Simultaneously, the cloud services platform can identify the recipient, and provide the customers with the query interface. The two parts of the system can communicates with each other in wireless channel through the communication unit.

Abstract: Near infrared spectroscopy is a promising technique in many scientific fields, but it is still a new measurement in sport science, of which coaches and athletes cannot make good use. In this paper we demonstrate how the real-time physiological oxygenated hemoglobin and deoxygenated hemoglobin concentration changes in the muscle can be measured noninvasively by near infrared spectroscopy and explain what the advantages of it is. The purpose is to reveal the mechanism and potential of muscle oxygen monitoring, which may play an important role in sport training.

Abstract: In order to solve the problem of complicated electronic equipment structure, inadequate fault information, hard to predict the fault and the existing failure prediction method cannot predict the state of the electronic equipment and other issues directly, we propose a combination failure prediction methods of least squares support vector machine (LSSVM) and hidden Markov model (HMM) based on Condition Based Maintenance (CBM). First, according to sensitivity analysis to determine the circuit elements to be changed to set the circuit by changing the parameters of the different components degraded state; secondly, create a combination failure prediction model; Finally, the circuit state prediction. The results show that the proposed method can directly predict the different states of the circuit, so as to realize the fault state prediction of the electronic equipment directly, the prediction accuracy can reach 93.3%.

Abstract: The paper using electrical capacitance tomography to test the flow velocity and flow rate of the mud transporting in the dredging.The testing process based on two 12 electrodes capacitive sensor sets (ie,the upstream and downstream sensors).The computer can gather the capacitance signals from the upstream and downstream sensor,then by the method of cross-correlation method it can calculate the flow velocity and flow rate of the mud in the slurry pipelines.In oader to verify the measurement error,the system measure the flow velocity and flow rate of the mud in the same situation at the different time (including five different time).Comparing the results of related calculations and the actual measurement it can prove that the cross-relation method is feasible and reliable,the tesing result have high accuracy.

Abstract: With the development of information and communication technology and the advent of new technologies including wireless communication, smart handheld mobile terminals and wearable medical sensor detection technology, mobile medical care is facing its development opportunity. In this context, our medical monitoring system combines wireless communication technology with wearable medical sensor and designs a remote medical monitoring of multiple physiological parameters. With the support of Android devices, this system offers medical reports retrieval, online diagnosis and health guidelines. It provides users with multi-functional, instant, real-time monitoring and interactive medical services in a high efficient low cost way, which enables remote interrogation and medical care become possible.

Abstract: Classing object is an important step for the high-level visions processing tasks, such as security managing, and abnormality event analysis. In this paper, we address these challenges of abnormal water surface monitoring in real-world unconstrained environments where the background is complex and dynamic. In the algorithm proposed, we extract the moment features of water surface in HSI space, and a technique is developed to monitor the abnormal surface of water based on moment features. Experimental results show that our algorithm works efficiently and robustly.