A5005550458- Industrial Vision Systems and Defect Detection
- Infrared Target Detection Methodologies
- Infrastructure Maintenance and Monitoring
- Surface Roughness and Optical Measurements
- Optical Polarization and Ellipsometry
- Thermography and Photoacoustic Techniques
- Image Enhancement Techniques
- Advanced Neural Network Applications
- Advanced Image Fusion Techniques
- Advanced SAR Imaging Techniques
- Visual perception and processing mechanisms
- Welding Techniques and Residual Stresses
- Video Surveillance and Tracking Methods
- Infrared Thermography in Medicine
- Advanced Measurement and Detection Methods
- Optical Systems and Laser Technology
- Astronomical Observations and Instrumentation
- Inertial Sensor and Navigation
- Neural dynamics and brain function
Nanjing University of Aeronautics and Astronautics
2013-2025
Ministry of Industry and Information Technology
2020-2022
The quality inspection of solder joints on aviation plugs is extremely important in modern manufacturing industries. However, this task still mostly performed by skilled workers after welding operations, posing the problems subjective judgment and low efficiency. To address these issues, an accurate automated detection system using fine-tuned YOLOv5 models developed paper. Firstly, we design intelligent image acquisition to obtain high-resolution each joint automatically. Then, a two-phase...
Surface defect detection (SDDet) is extremely critical for quality control and routine maintenance, computer vision-based methods have delivered promising performance in various industrial fields. However, there remains big challenges region consistency boundary localization, due to complicated appearance defects, interference of artifacts, low contrast. To overcome these issues, this article, we propose an enhanced encoder–decoder network with hierarchical supervision SDDet. Specifically,...
Surface defects have an adverse effect on the quality of industrial products, and vision-based defect detection is widely researched due to its objective stable performance. However, task still challenging diversified types complex background texture. The robust principal component analysis (RPCA) has proven applicable in inspection by regarding nondefective as low-rank part defective area sparse part. such methods cannot sufficiently detect cluttered background, noise interference, limited...
The responses of cortical neurons to a stimulus in classical receptive field (CRF) can be modulated by stimulating the non-CRF (nCRF) primary visual cortex (V1). In very early stages (at around 40 ms), neuron V1 exhibits strong small set stimuli. Later, however (after 100 become sensitive scene's global organization. As per these mechanisms, contour detection model based on spatial summation properties is proposed. Unlike previous studies, nCRF higher that results inhibition neuronal...
For infrared (IR) small target detection, we propose a new spatial-temporal tensor (STT) decomposition model based on the Schatten capped <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$p$ </tex-math></inline-formula> norm (TSC notation="LaTeX">$p\text{N}$ ) and total variation (TV) regularization. First, to explore spatial temporal information, construct an STT introduce prior weight map. Then, replacing...
For the purpose of high-accuracy detection and tracking sun position, a measuring sensor based on Position Sensitive Detectors(PSD)has been designed. The PSD is installed at bottom canister which has narrow light pinhole so that could pass through it. Sunlight spot projects surface then information about position can be captured PSD. solar changed into electrical signal by processing circuit, further send to follow-up system. experimental results show accuracy this 0.05°at ±12°acceptable...
Abstract Infrared small target detection plays an important role in infrared search and tracking applications. In order to solve the problem of has high false alarm rates low probabilities with complex backgrounds. An method based on local contrast measure gradient property (LCG) is proposed this paper. Initially, map input image calculated using modified which employed enhance suppress background. Then, according distribution regularly around target, from further clutters. Next, feature...