Mr. Lu

Prof. Dr. Mr. Lu


Research Focus & Metrics

Academic Career Length

21Years (2006-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 May 2023 in Engineering Applications of Artificial Intelligence
Verified
Journal Article
Closed access
DWDahang Wan
Dahang Wan
MLMr. Lu
Mr. Lu
SSSiyuan Shen
Siyuan Shen
TXTing Xu
Ting Xu
XLXianli Lang
Xianli Lang
ZRZhijie Ren
Zhijie Ren

8 August 2011 in Sensors
Verified
Journal Article
Open access
PDF file
YZYongkai Zhu
Yongkai Zhu
Gui-Yun TianGui-Yun Tian
Gui-Yun Tian
MLMr. Lu
Mr. Lu
HZHong Zhang
Hong Zhang
Optical non-destructive testing (NDT) has gained more and more attention in recent years, mainly because of its non-destructive imaging characteristics with high precision and sensitivity. This paper provides a review of the main optical NDT technologies, including fibre optics, electronic speckle, infrared thermography, endoscopic and terahertz technology. Among them, fibre optics features easy integration and embedding, electronic speckle focuses on whole-field high precision detection, infrared thermography has unique advantages for tests of combined materials, endoscopic technology provides images of the internal surface of the object directly, and terahertz technology opens a new direction of internal NDT because of its excellent penetration capability to most of non-metallic materials. Typical engineering applications of these technologies are illustrated, with a brief introduction of the history and discussion of recent progress.

20 January 2023 in Remote Sensing
Verified
Journal Article
Open access
PDF file
DWDahang Wan
Dahang Wan
MLMr. Lu
Mr. Lu
SWSailei Wang
Sailei Wang
SSSiyuan Shen
Siyuan Shen
TXTing Xu
Ting Xu
XLXianli Lang
Xianli Lang
Object detection is essential to the interpretation of optical remote sensing images and can serve as a foundation for research into additional visual tasks that utilize remote sensing. However, the object detection network currently employed in optical remote sensing images underutilizes the output of the feature pyramid, so there remains potential for an improved detection. At present, a suitable balance between the detection efficiency and detection effect is difficult to attain. This paper proposes an enhanced YOLOv5 algorithm for object detection in high-resolution optical remote sensing images, utilizing multiple layers of the feature pyramid, a multi-detection-head strategy, and a hybrid attention module to improve the effect of object-detection networks for use with optical remote sensing images. According to the SIMD dataset, the mAP of the proposed method was 2.2% better than YOLOv5 and 8.48% better than YOLOX, achieving an improved balance between the detection effect and speed.

10 December 2006 in Applied Optics
Verified
Journal Article
Closed access
PDF file
MLMr. Lu
Mr. Lu
GTGui-Yun Tian
Gui-Yun Tian
DGDuke Gledhill
Duke Gledhill
SWSteve Ward
Steve Ward
Surface speckle pattern intensity distribution resulting from laser light scattering from a rough surface contains various information about the surface geometrical and physical properties. A surface roughness measurement technique based on the texture analysis of surface speckle pattern texture images is put forward. In the surface roughness measurement technique, the speckle pattern texture images are taken by a simple setup configuration consisting of a laser and a CCD camera. Our experimental results show that the surface roughness contained in the surface speckle pattern texture images has a good monotonic relationship with their energy feature of the gray-level co-occurrence matrices. After the measurement system is calibrated by a standard surface roughness specimen, the surface roughness of the object surface composed of the same material and machined by the same method as the standard specimen surface can be evaluated from a single speckle pattern texture image. The robustness of the characterization of speckle pattern texture for surface roughness is also discussed. Thus the surface roughness measurement technique can be used for an in-process surface measurement.

1 May 2023 in Scientific Reports
Verified
Journal Article
Open access
PDF file
JWJyunrong Wang
Jyunrong Wang
HDHuafeng Dai
Huafeng Dai
TCTaogen Chen
Taogen Chen
HLHao Liu
Hao Liu
XZXuegang Zhang
Xuegang Zhang
MLMr. Lu
Mr. Lu
In electronics manufacturing, surface defect detection is very important for product quality control, and defective products can cause severe customer complaints. At the same time, in the manufacturing process, the cycle time of each product is usually very short. Furthermore, high-resolution input images from high-resolution industrial cameras are necessary to meet the requirements for high quality control standards. Hence, how to design an accurate object detector with real-time inference speed that can accept high-resolution input is an important task. In this work, an accurate YOLO-style object detector was designed, ATT-YOLO, which uses only one self-attention module, many-scale feature extraction and integration in the backbone and feature pyramid, and an improved auto-anchor design to address this problem. There are few datasets for surface detection in electronics manufacturing. Hence, we curated a dataset consisting of 14,478 laptop surface defects, on which ATT-YOLO achieved 92.8% mAP0.5 for the binary-class object detection task. We also further verified our design on the COCO benchmark dataset. Considering both computation costs and the performance of object detectors, ATT-YOLO outperforms several state-of-the-art and lightweight object detectors on the COCO dataset. It achieves a 44.9% mAP score and 21.8 GFLOPs, which is better than the compared models including YOLOv8-small (44.9%, 28.6G), YOLOv7-tiny-SiLU (38.7%, 13.8G), YOLOv6-small (43.1%, 44.2G), pp-YOLOE-small (42.7%, 17.4G), YOLOX-small (39.6%, 26.8G), and YOLOv5-small (36.7%, 17.2G). We hope that this work can serve as a useful reference for the utilization of attention-based networks in real-world situations.

1 April 2011 in Optical Engineering
Verified
Journal Article
Closed access
PDF file
JDJing-Tao Dong
Jing-Tao Dong
MLMr. Lu
Mr. Lu
YSYan-Qiong Shi
Yan-Qiong Shi
RXRui-Xue Xia
Rui-Xue Xia
QLQi Li
Qi Li
YXYan Xu
Yan Xu
Uniform irradiance and color adjustability are the key features in the design of lighting for machine vision inspection systems. A simple and practical design tool of angled light-emitting-diode (LED) ring arrays for uniform near-field irradiance has been developed by introducing a simple model to simplify the complexity of nonrotational symmetric irradiance distribution of angled LEDs. The color distribution and color uniformity of the ring array assembled with RGB LEDs are analyzed based on the analytical model of color mixing. According to the theoretical analysis, the simulated results, and the design exemplifications, the practical design tool offers an easy way to estimate the performance of an RGB LED ring array and can be considered as a starting point to reduce the computation time for exact designs that must use a realistic LED model.