A5019721709- Remote Sensing and Land Use
- Advanced Algorithms and Applications
- Infrared Target Detection Methodologies
- Power Line Inspection Robots
- Advanced Measurement and Detection Methods
- Spectroscopy and Chemometric Analyses
- Traffic Prediction and Management Techniques
- Smart Agriculture and AI
- Stability and Control of Uncertain Systems
- Acoustic Wave Resonator Technologies
- Wood and Agarwood Research
- Safety and Risk Management
- QR Code Applications and Technologies
- Mechanical and Optical Resonators
- Zebrafish Biomedical Research Applications
- Fuzzy Logic and Control Systems
- Soft Robotics and Applications
- Indoor and Outdoor Localization Technologies
- IPv6, Mobility, Handover, Networks, Security
- Face recognition and analysis
- Advanced MEMS and NEMS Technologies
- Remote-Sensing Image Classification
- Caching and Content Delivery
- Adaptive Dynamic Programming Control
- Neurological disorders and treatments
Harbin University of Science and Technology
2011-2024
Lamar University
2002
The detection of wood defects plays an important role in the processing and production wood, deep learning methods have achieved outstanding results this field recent years. There are still two unresolved problems, one is that category unbalanced difficult to obtain, other effective modeling with different sizes irregular shapes. Here alternative method proposed. For problem imbalance defect categories, a generative data augmentation method, cycle adversarial network (GAN), used generate new...
Objective: To explore the application value of artificial intelligence-assisted diagnosis system for TBS report in cervical cancer screening. Methods: A total 16 317 clinical samples and related data liquid-based thin-layer cell smears, which were obtained from July 2020 to September 2020, collected Southern Hospital, Guangzhou Huayin Medical Inspection Center, Shenzhen Bao'an People's Hospital(Group) Changsha Yuan'an Biotechnology Co., Ltd. The cytology jointly developed by University F. Q....
Aiming at the low efficiency and manufacturing flexibility of filament winding machine, a control system based on robot that controls motion trajectory silking mouth for elbow pipe is proposed in this paper. The traditional method either manual operation or semi-automated, whose accuracy are low. Filament machine has many advantages, such as flexibility, multiple degrees freedom, wide range process. This paper presents overall design hardware selection system, introduces stable path planning robot.
The Color sorting technology is the core of color sorters. At present there are many image processing and recognition techniques applied to technology. In this paper, we Image Texture Analysis methods identify quality rice. We can collect rice through CCD camera. Using GLCM (Gray Level Co - occurrence Matrix) algorithm deal with acquisition data image, by calculating Energy, Entropy, Moment Inertia Relevance characteristic values at 0°, 45°, 90°, 135° direction. And comparing standard intact...
Multi-tag anti-collision algorithm is becoming not only a focus in the research field of RFID key technology but also one most important topics for Internet Things. Therefore, dynamic frame slotted ALOHA has been found to specialize problem anti-collision. Comparing with pure ALOHA, algorithm, obvious advantages. The conditions and realization methods maximum throughput rate, which acquires, have obtained through analysis simulation.
Based on BP neural network to control the complex hydraulic gap (HGC) system and point out boundness of selection uncertainty for layers neurons randomness connection weights between layers. In this paper, an improved PID (PIDNN) is proposed make trapezoidal integral transform hidden neuron nodes incomplete differential transformation nodes. The output function each node hyperbolic tangent replace proportion threshold function. To by PIDNN, simulation results show that has better efficiency...
A taxonomy for coordinated computational intelligence (CCT) is introduced fuzzy-neural control. Based on the CCI a multiagent cerebellar model MAC-J, extended from 3-link uniped locomotion training to 4-link and usual robot gymnastic training. Basic ideas are illustrated with examples examples. Implications of this work human animal control its potential applications discussed.
Using color linear array CCD (Charge Coupled Device) to collect the information of image has an important meaning in research machine vision processing. It now becomes focus study on acquisition using FPGA (Field Programmable Gate Array) implement CCD-driven design. The and processing rice sorter as background this paper, by achieve driven control, well simulated with Quartus II 8.0. In order be able extract output useful images, through RGB model HSI (Huge, Saturation, Intensity) space...
On the basis of Internet communication, this paper puts forward a kind combination virtual private network VPN technology equipment remote monitoring solutions. In client build Server services and use SSL C/S mode structure, control access permissions, protect data security transfer. Expound general structure design gateway. series cryptogram knowledge related, including symmetric encryption, asymmetric digital signatures, certificates message digest,this careful analysis principles relevant...
Linear array CCD (Charge Coupled Device) image sensor has a lot advantage compared to CMOS (Complementary Metal Oxide Semiconductor) be widely used. Recently, it been paid more attention that FPGA (Field Programmable Gate Array) is applied the preprocessing field. In this paper, rice color sorter study object. aspect of hardware design, high-speed acquisition based on linear designed with and simulated. software fast uniform motion restoration algorithm solve blurred problem caused by...
Magnetic resonance imaging (MRI)-based electroencephalography (EEG) forward modeling method has become prevalent in the field of EEG. However, due to inability obtain clear images an infant's fontanel through MRI, fontanelle information is often lacking EEG model, which affects accuracy infants. To address this issue, we propose a novel achieve compensation for infant method. First, employed segmentation and meshing head MRIs, creating fontanel-free model. Second, projection-based surface...
Y-cut quartz resonators operating on low-frequency thickness C-mode is chosen as the temperature sensor because of its availability and good linear frequency versus characteristic. In order to reduce mounting effects decouple interfering modes motion, contour structures wafers having a biconvex type bat metallic electrodes with sine ripple surfaces wafer are applied. The resonance characteristic analyzed by finite element method results compared theoretical values whenever they available....
Multi-sensor information fusion technology can effectively reduce the risk of car theft and false alarm rate. This paper proposes multi-sensor in vehicle anti-theft system based on improved D-S evidence theory. The use method with theory process conflict reasonably, so that is closer to real situation. When hurt it realize real-time control through GSM network, solving problem traditional effectively.
In accordance with the characteristics of paddy images, this article designs a rice image acquisition device which consisted one low-cost CCD sensor camera, two LED lamps, detection box, test-bed, and PC computer. The was designed to work facilitate study processing algorithms. Firstly, algorithm wavelet packet mathematical morphology in method is taken smooth interferences surface scratches glistens effectively. Secondly, processed detected edge Canny operator. paper, we used three...
Intelligent transportation systems (ITS) have developed for a long time. The rise of deep learning has brought new vitality the ITS. However, traffic flow data is usually time-correlated and highly randomized. distribution will also change dynamically. To actualize forecasting accurately, we use historical information to predict messages at any time interval. This paper proposes an efficient forecast architecture based on learning. method combines gated recurrent unit (GRU, type neural...
Rice is the common food in people's daily life, different grades of rice have a great difference to influence taste, therefore, it important identify quality. In order improve accuracy and real-time performance quality identification, PCA algorithm (Principal Component Analysis) BP neural network are proposed The feature variables extracted by algorithm, then training parameters with as its input, finally trained used implemented into MATLAB software. simulation shows that PCA-BP greatly...
A classification algorithm was developed to differentiate individual infected (dead, chalky, cracked, and immature) qualified rice kernels. The image preprocessed by wavelet packet, the feature regions of interest were extracted edge detection. Ten statistical features (area, perimeter, compactness, etc.) from data single composed pattern vector a kernel. dimensionality vectors reduced principal component analysis. multi-class support machine with kernel radial basis function used for...
Aiming at the limitations of clinical diagnosis Parkinson's disease (PD) with rapid eye movement sleep behavior disorder (RBD), in order to improve accuracy diagnosis, an intelligent-aided method based on few-channel electroencephalogram (EEG) and time-frequency deep network is proposed for PD RBD. Firstly, speed operation robustness algorithm, 6-channel scalp EEG each subject were segmented same time-window. Secondly, model was constructed trained time-window data obtain segmentation-based...
Color is a primary way of people perception for recognizing image, with the development science and technology improvement computer processing capabilities, it possible researchers to deal multidimensional image data in color space, regarded as new hotspot processing. Many edge detection algorithms grayscale images are applied divide detection; however, without making use information effective, this kind cannot get reasonable result. By studying characteristics HSI discussing its H (Huge)...