Autonomous planning of navigation trajectory is a key research element of USV and a basic condition for safe and autonomous navigation. In this paper, an autonomous planning collision avoidance control algorithm for USV is introduced. TFMS is used to trajectory planning and dynamic window method for real-time collision avoidance, a hybrid planning algorithm of dynamic path planning algorithm and parallel planning of autonomous collision avoidance control algorithm is built to achieve autonomous navigation control of USV. The setting method of local target points has been improved to shorten the running time of the control system. The validation was carried out in different static and dynamic environments. A realistic collision avoidance environment model is established by extracting the environmental data from the port map of Dalian port. The result demonstrates that the method is effective in controlling the USV to sail safely and quickly in complex real scenarios with various dynamic impediments.
Human activity recognition (HAR) has become a research hotspot in the field of artificial intelligence and pattern recognition. However, the HAR system still has some deficiencies in the aspects of platform algorithms and wireless access technologies. On the one hand, some state-of-the-art frameworks such as convolutional neural network (CNN) and recurrent neural network (RNN) have been proven successfully in classification tasks of HAR, while those frameworks just identify the feature data of activity but ignore the spatial relationship among features, which may lead to incorrect recognition. On the other hand, some existing transmission modes, such as Bluetooth and 4G, are difficult to realize real-time transmission in the case of a large range and low-power consumption. In this paper, a real-time human activity recognition system based on capsule and “long range” (LoRa) is presented, which pioneers the application of capsule to HAR. The capsule framework encapsulates the multiple convolution layers in parallel to solve the defect that current frameworks cannot identify the spatial relationship among features. Simultaneously, the combination of long-distance transmission and low-power consumption is achieved by using LoRa networking technology instead of other existing transmission modes. The experiments are performed on the dataset WISDM that is collected by the Wireless Sensor Data Mining Lab in Fordham University, and the results demonstrate that the proposed capsule framework achieves a higher classification result than CNN and RNN, and the proposed system makes the real-time HAR based on intelligent sensor devices possible in some special scenarios such as smart prison.
In this paper, a method of face recognition based on shearlet transform and fast independent component analysis (Fast ICA) is proposed to overcome the disadvantage of shearlet transform, which is easy to have data redundance in extracting features. First of all, the coefficients of different scales and directional sub bands are obtained after using shearlet transform to the face images, then using Fast ICA for further extraction to eliminate the high-level redundancy. At last, using support vector machine for classification. In ORL face databases, the experimental results show that the algorithm has a high recognition performance and can capture the facial features effectively.
The Spectral Image Decomposition (SPID) compres- sor uses techniques borrowed from spectral image analysis to achieve a high degree of compression on hyperspectral images. The purpose of this compressor is to provide real-time compres- sion for images captured one line at a time using a push-broom method. In addition, the process is aimed at platforms such as satellites where only limited hardware may be deployed; memory, CPU time, and storage space must all be conserved. Spectral image decomposition and temporal differencing techniques were used to achieve a compression ratio of 23:1 in a low error near- lossless mode and 4.25:1 in lossless mode. SPID is a flexible compressor that can provide optimal performance for a variety of objectives.
The generation of three-dimensional tunable vector optical cages through full polarization modulation requires complex polarization states. This paper takes the vector Airy optical cage as an example to generate a three-dimensional tunable high-quality optical cage based on the Pancharatnam-Berry phase principle. The proposed method in this paper possesses the capability of arbitrary modulation in various aspects, including the quantity of optical cages and their respective sizes as well as three-dimensional spatial positions. Moreover, the intensity of each optical cage can be modulated independently. This research will improve the capture efficiency of optical tweezers and promote further development in fields of efficient optical trapping, particle manipulation, high-resolution microscopic manipulation, and optical communication.
In this paper, by using a pinning adaptive control scheme, we investigate the almost surely synchronization of neutral-type coupled neural networks with stochastic perturbation. Based on Lyapunov stability theory, stochastic analysis, and matrix theory, some sufficient conditions for almost surely synchronization are derived. Furthermore, a numerical example is exhibited to illustrate the validity of the theoretical results.
The advances in deep learning with the ability to automatically extract advanced features have achieved a bright prospect for human activity recognition (HAR). However, the traditional HAR methods still have the deficiencies of incomplete feature extraction, which may lead to incorrect recognition results. To resolve the above problem, a novel framework for spatiotemporal multi-feature extraction on HAR called CapsGaNet is propounded, which is based on capsule and gated recurrent units (GRU) with attention mechanisms. The proposed framework involves a spatial feature extraction layer consisting of capsule blocks, a temporal feature extraction layer consisting of GRU with attention mechanisms, and an output layer. At the same time, considering the actual demands for recognizing aggressive activities in some specific scenarios like smart prison, we constructed a daily and aggressive activity dataset (DAAD). Moreover, based on the acceleration characteristics of aggressive activity, a threshold-based approach for aggressive activity detection is propounded to meet the needs of high real-time and low computational complexity in prison scenarios. The experiments are performed on the wireless sensor data mining (WISDM) dataset and the DAAD dataset, and the results verify that the propounded CapsGaNet could effectually improve the recognition accuracy. The proposed threshold-based approach for aggressive activity detection provides a more effective HAR way by using smart sensor devices in smart prison scenarios.
As rail vehicles cover more mileage and run at higher speeds, the problem of abnormal wear between the wheel and rail becomes increasingly prominent. Identifying wheel polygons is a crucial aspect of tackling this issue. In this study, acceleration signals are collected using the wheel rotation angle as the reference instead of time, and signal processing is carried out using the spatial spectrum Fourier transform. The results demonstrate that the Fourier transform based on the spatial spectrum can accurately identify the abnormal frequency of polygon distribution. This method solves the problem of polygon order recognition under changing train speeds and maintains good recognition accuracy. Simulation results show that the recognition accuracy under different speed segments is higher than 95%, and polygon order recognition can be conducted under different working conditions with robustness.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
