Automatic laser profile recognition and fast tracking for structured light measurement using deep learning and template matching

Wang, Shengchun; Wang, Hao; Zhou, Yunlai; Liu, Junbo; Dai, Peng; Du, Xinyu; Wahab, Magd Abdel

doi:10.1016/j.measurement.2020.108362

Cited by 114 publications

(21 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, they may suffer from robustness problems due to complex backgrounds. Wang et al [6] combined a deep learning model and a template-matching driven tracking algorithm for recognition and tracking of rail profile from laser fringe images. Jiang et al [7] proposed a robust line detection workflow for the uplift measurement of railway catenary, addressing the problem caused by noisy background.…”

Section: Background and Motivationmentioning

confidence: 99%

Deep learning based virtual point tracking for real-time target-less dynamic displacement measurement in railway applications

Shi

Šabanovič

Rizzetto

et al. 2022

Mechanical Systems and Signal Processing

View full text Add to dashboard Cite

In the application of computer-vision-based displacement measurement, an optical target is usually required to prove the reference. If the optical target cannot be attached to the measuring objective, edge detection and template matching are the most common approaches in target-less photogrammetry. However, their performance significantly relies on parameter settings. This becomes problematic in dynamic scenes where complicated background texture exists and varies over time. We propose virtual point tracking for real-time target-less dynamic displacement measurement, incorporating deep learning techniques and domain knowledge to tackle this issue. Our approach consists of three steps: 1) automatic calibration for detection of region of interest; 2) virtual point detection for each video frame using deep convolutional neural network; 3) domain-knowledge based rule engine for point tracking in adjacent frames. The proposed approach can be executed on an edge computer in a real-time manner (i.e. over 30 frames per second). We demonstrate our approach for a railway application, where the lateral displacement of the wheel on the rail is measured during operation. The numerical experiments have been performed to evaluate our approach's performance and latency in a harsh railway environment with dynamic complex backgrounds. We make our code and data available at https://github. com/quickhdsdc/Point-Tracking-for-Displacement-Measurement-in-Railway-Applications.

show abstract

Section: Background and Motivationmentioning

confidence: 99%

Deep learning based virtual point tracking for real-time target-less dynamic displacement measurement in railway applications

Shi

Šabanovič

Rizzetto

et al. 2022

Mechanical Systems and Signal Processing

View full text Add to dashboard Cite

show abstract

“…Machine vision is the main technology used to achieve high detection accuracy for the identification of an object and its orientation. In the literature, the most commonly applied methods for object detection have been template matching methods (25)(26)(27)(28) and feature-point matching methods. (29,30) Because the shaft nut on the spindle of the round bar sawing equipment is hexagonal and symmetric, the developed template matching approach is applied in this work.…”

Section: Spindle Location Detection Via Template Matching Approachmentioning

confidence: 99%

Machine-vision-based Spindle Positioning System of Grinding-wheel-saw Automatic Replacement System

Qiu¹,

Yang²,

Wu³

et al. 2022

Sensors and Materials

View full text Add to dashboard Cite

To improve the working efficiency and safety of steel round bar sawing equipment, a grinding-wheel-saw automatic replacement system has been employed to replace grinding-wheel saws. In this paper, we report a machine-vision-based spindle positioning system for spindle localization in this replacement system. The hardware and software for the spindle positioning system were established, as well as an imaging model of the camera and the calibration of the vision system. The template matching approach applied to identify the shaft nut of the spindle is also described. Experimental results clearly demonstrate the effectiveness of the proposed machine-vision-based system.

show abstract

“…According to Equation (15), when the object distance is z Q and the UAV carrying the camera with focal length f and baseline B, in order to reduce the distance error ∆z Q , it can be achieved by increasing the baseline B. When the baseline is 2.5 m, the UAV positioning error ∆B is 1 mm, and the object distance z Q is 3 m, the measurement error ∆z Q can be controlled to 1.2 mm.…”

Section: Controlling Precision Analysismentioning

confidence: 99%

“…The structured light measurement system is widely applicable to various fields of industrial measurement. Wang et al [15] proposed rail profile recognition based on structured light measurement with depth learning and template matching. The advantages of structured light 3D shape measurement are high precision and high resolution, but the measurable size is limited [16].…”

Section: Introductionmentioning

confidence: 99%

LiDAR-Assisted UAV Stereo Vision Detection in Railway Freight Transport Measurement

Zhou

Gong

et al. 2022

Drones

View full text Add to dashboard Cite

Identifying and detecting the loading size of heavy-duty railway freight cars is crucial in modern railway freight transportation. Due to contactless and high-precision characteristics, light detection and ranging-assisted unmanned aerial vehicle stereo vision detection is significant for ensuring out-of-gauge freight transportation security. However, the precision of unmanned aerial vehicle flight altitude control and feature point mismatch significantly impact stereo matching, thus affecting the accuracy of railway freight measurement. In this regard, the altitude holding control strategy equipped with a laser sensor and SURF_rBRIEF image feature extraction and matching algorithm are proposed in this article for railway freight car loading size measurement. Moreover, an image segmentation technique is used to quickly locate and dismantle critical parts of freight cars to achieve a rapid 2-dimension reconstruction of freight car contours and out-of-gauge detection. The robustness of stereo matching has been demonstrated by external field experiment. The precision analysis and fast out-of-gauge judgment confirm the measurement accuracy and applicability.

show abstract

Automatic laser profile recognition and fast tracking for structured light measurement using deep learning and template matching

Cited by 114 publications

References 14 publications

Deep learning based virtual point tracking for real-time target-less dynamic displacement measurement in railway applications

Deep learning based virtual point tracking for real-time target-less dynamic displacement measurement in railway applications

Machine-vision-based Spindle Positioning System of Grinding-wheel-saw Automatic Replacement System

LiDAR-Assisted UAV Stereo Vision Detection in Railway Freight Transport Measurement

Contact Info

Product

Resources

About