Search

As large-volume metrology gains significance in industrial applications, ensuring high-accuracy measurements (±0.1 mm) with cost-effective and user-friendly technologies becomes increasingly challenging. This study evaluates the performance of Markerless Photogrammetry (MP) against industry standards such as Coordinate Measuring Machines (CMM), Laser Trackers (LT), and Marker-based Photogrammetry (MbP) to identify the potential of MP in addressing industrial challenges. While CMM and LT offer high accuracy, their applicability is limited by the size of large workpieces and cost. MbP offers high work ranges, flexibility and ease of use, but it is restricted to sparse measurements and lacks the accuracy of traditional methods. Despite its lower accuracy, the results indicate that MP may present a viable solution for cases where cost-effectiveness and ease of use are crucial with the drawback of high computational cost.

Currently, the rapid development of the industrial Internet has led to the creation of a massive number of intelligent agents that are widely and distributively applied in various edge scenarios. The work conditions in these edge scenarios are complex, uncertain, and random. Traditional manual updates or human judgments are used for task decision-making in large-scale distributive intelligent agent edge work scenarios, which lack dynamic perception and autonomous recognition capabilities for edge work conditions. This inevitably leads to low decision-making accuracy, poor reliability, and ultimately, task failure. To address this issue, this study proposes an adaptive task identification strategy based on cloud-edge collaboration. This method utilizes a cloud-edge collaborative industrial intelligent application architecture to achieve cloud-based training and encapsulation of the task model, with online calling at the edge-end. Then, edge-end intelligent agents identify edge work conditions through multi-source data fusion, enabling accurate task decision-making. Finally, the edge-end requests the cloud for task model matching. The effectiveness of the proposed method is validated in an industrial safety situation virtual detection system.

Currently, the rapid development of the industrial Internet has led to the creation of a massive number of intelligent agents that are widely and distributively applied in various edge scenarios. The work conditions in these edge scenarios are complex, uncertain, and random. Traditional manual updates or human judgments are used for task decision-making in large-scale distributive intelligent agent edge work scenarios, which lack dynamic perception and autonomous recognition capabilities for edge work conditions. This inevitably leads to low decision-making accuracy, poor reliability, and ultimately, task failure. To address this issue, this study proposes an adaptive task identification strategy based on cloud-edge collaboration. This method utilizes a cloud-edge collaborative industrial intelligent application architecture to achieve cloud-based training and encapsulation of the task model, with online calling at the edge-end. Then, edge-end intelligent agents identify edge work conditions through multi-source data fusion, enabling accurate task decision-making. Finally, the edge-end requests the cloud for task model matching. The effectiveness of the proposed method is validated in an industrial safety situation virtual detection system. [ABSTRACT FROM AUTHOR]

In modern industrial production, high-temperature environments are commonplace, posing significant challenges to equipment stability, safety, and production efficiency. Machine vision, as an effective automated inspection technology, has attracted extensive attention in high-temperature settings. However, the unique conditions of high temperatures, such as significant thermal noise and optical interference, demand enhanced performance from machine vision systems. The second law of thermodynamics provides a theoretical foundation for understanding these challenges, emphasizing the increase of entropy in energy transformation and transfer processes, and guides the design and optimization of machine vision systems in high-temperature environments. This paper aims to comprehensively explore the application of machine vision based on the second law of thermodynamics in high-temperature industrial inspection, focusing on two core issues: the impact of thermodynamic parameters on the performance of machine vision systems and the technology for analyzing high-temperature industrial infrared images using multiscale entropy. By thoroughly analyzing how thermodynamic parameters influence the design and implementation of machine vision systems, and by developing infrared image processing algorithms adapted to high temperatures, this study seeks to enhance the efficiency and accuracy of machine vision technology in high-temperature industrial applications, providing theoretical support and technical guidance for the advancement of intelligent manufacturing. [ABSTRACT FROM AUTHOR]

In oil and gas offshore platforms, special pipelines as flexible risers make the connection between the ocean floor structures and the platform in extreme environmental and operational conditions. Periodic inspections are necessary to assess their integrity. As industrial climbing for inspection is expensive, extremely dangerous and time consuming, qualitative visual inspection with Remotely Piloted Aircraft System (RPAS), also known as drones, are being successfully applied for remote inspection of offshore flares and risers in a much safer, quicker, and cheaper way. These experiences motivate the 3D photogrammetric inspection of risers using RPAS, considering restrictions like layout of the inspected structures and surroundings and inability to prepare the scene. In this paper, taking advantage of the position information provided by the RPAS, the reconstruction and scale of the test scene were made using only GNSS data, GNSS and scale bars, RTK, and RTK and scale bars. Calibrated artifacts were used to evaluate the results and they include a PVC pipe with artificial defects simulating a riser, a pyramidal pattern with four spheres, and scale bars. The results showed that, as expected, the worst results are for GNSS data with error standard deviations of 0.35 mm compared with 0.20 mm or less for other options. For the sphere’s artifact, relative maximum sphere spacing errors are 9.3% for GNSS, 1.9% for RTK and 0.26% using scale bars. In any case it was possible to identify the defects in the pipe with good quality and with much more detail compared with a climbing inspection.

The weld defects are formed due to the incorrect welding patterns or wrong welding process. The defects in the weld may vary from size, shape and their projected quality. The most common weld defects occur during welding process is slag inclusions, porosity, lack of fusion and incomplete penetration. In this study, an effective method for weld defect classification using machine learning algorithm is presented. The system uses Speeded-up Robust Features (SURF) for feature extraction and one of the machine learning algorithms called Auto-Encoder Classifier (AEC) for classification. Initially, the features that distinguish weld defects and no defects in the weld image are extracted by SURF. Then, AEC is analyzed for weld image classification using different number of neurons in different hidden layers (2 and 3 hidden layers). The performance of the system is evaluated by GD X-ray weld image database. The results show that the weld images are correctly classified with 98% accuracy using SURF and AEC.

This study presents a complete solution for multi-modal inspection of industrial components, including a processing pipeline for registering consecutive multi-modal point clouds comprising thermal and visible sensors' data. A comparative evaluation of optimisation and learning-based registration methods is provided as part of the processing pipeline. Moreover, a benchmark dataset of point cloud data from different FOVs of industrial and construction component samples is provided (LeManchot-Points), having data from five point clouds with depth, colour and thermal information at each point. The experimental campaign with different objects demonstrates the proposed solution's applicability for the multi-modal inspection of industrial components. [ABSTRACT FROM AUTHOR]

As large-volume metrology gains significance in industrial applications, ensuring high-accuracy measurements (±0.1 mm) with cost-effective and user-friendly technologies becomes increasingly challenging. This study evaluates the performance of Markerless Photogrammetry (MP) against industry standards such as Coordinate Measuring Machines (CMM), Laser Trackers (LT), and Marker-based Photogrammetry (MbP) to identify the potential of MP in addressing industrial challenges. While CMM and LT offer high accuracy, their applicability is limited by the size of large workpieces and cost. MbP offers high work ranges, flexibility and ease of use, but it is restricted to sparse measurements and lacks the accuracy of traditional methods. Despite its lower accuracy, the results indicate that MP may present a viable solution for cases where cost-effectiveness and ease of use are crucial with the drawback of high computational cost. [ABSTRACT FROM AUTHOR]

An application of spectral domain optical coherence tomography (SD-OCT) was demonstrated for a fast industrial inspection of an optical thin film panel. An optical thin film sample similar to a liquid crystal display (LCD) panel was examined. Two identical SD-OCT systems were utilized for parallel scanning of a complete sample in half time. Dual OCT inspection heads were utilized for transverse (fast) scanning, while a stable linear motorized translational stage was used for lateral (slow) scanning. The cross-sectional and volumetric images of an optical thin film sample were acquired to detect the defects in glass and other layers that are difficult to observe using visual inspection methods. The rapid inspection enabled by this setup led to the early detection of product defects on the manufacturing line, resulting in a significant improvement in the quality assurance of industrial products.

Recently, researchers have been studying methods to introduce deep learning into automated optical inspection (AOI) systems to reduce labor costs. However, the integration of deep learning in the industry may encounter major challenges such as sample imbalance (defective products that only account for a small proportion). Therefore, in this study, an anomaly detection neural network, dual auto-encoder generative adversarial network (DAGAN), was developed to solve the problem of sample imbalance. With skip-connection and dual auto-encoder architecture, the proposed method exhibited excellent image reconstruction ability and training stability. Three datasets, namely public industrial detection training set, MVTec AD, with mobile phone screen glass and wood defect detection datasets, were used to verify the inspection ability of DAGAN. In addition, training with a limited amount of data was proposed to verify its detection ability. The results demonstrated that the areas under the curve (AUCs) of DAGAN were better than previous generative adversarial network-based anomaly detection models in 13 out of 17 categories in these datasets, especially in categories with high variability or noise. The maximum AUC improvement was 0.250 (toothbrush). Moreover, the proposed method exhibited better detection ability than the U-Net auto-encoder, which indicates the function of discriminator in this application. Furthermore, the proposed method had a high level of AUCs when using only a small amount of training data. DAGAN can significantly reduce the time and cost of collecting and labeling data when it is applied to industrial detection.

Copyright of Journal of Harbin Institute of Technology. Social Sciences Edition / Haerbin Gongye Daxue Xuebao. Shehui Kexue Ban is the property of Harbin Institute of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Anomaly detection is crucial in large-scale industrial manufacturing as it helps to detect and localize defective parts. Pre-training feature extractors on large-scale datasets is a popular approach for this task. Stringent data security, privacy regulations, high costs, and long acquisition time hinder the development of large-scale datasets for training and benchmarking. Despite recent work focusing primarily on the development of new anomaly detection methods based on such extractors, not much attention has been paid to the importance of the data used for pre-training. This study compares representative models pre-trained with fractal images against those pre-trained with ImageNet, without subsequent task-specific fine-tuning. We evaluated the performance of eleven state-of-the-art methods on MVTecAD, MVTec LOCO AD, and VisA, well-known benchmark datasets inspired by real-world industrial inspection scenarios. Further, we propose a novel method to create a dataset by combining the dynamically generated fractal images creating a “Multi-Formula” dataset. Even though pre-training with ImageNet leads to better results, fractals can achieve close performance to ImageNet under proper parametrization. This opens up the possibility for a new research direction where feature extractors could be trained on synthetically generated abstract datasets mitigating the ever-increasing demand for data in machine learning while circumventing privacy and security concerns.

3D object alignment is essential in multiple fields. For instance, to allow precise measurements in metrology, to perform surface/volumetric checks or quality control in industrial inspection, to align partial captures of a 3D object during object scanning, to simplify object recognition or classification in pattern recognition, accuracy and speed, being opposed, are desirable features of those algorithms. Nevertheless, they can be more or less critical depending on the application area. In the present work, we propose a methodology to improve the alignment of 3D objects reconstructed using shape-from-silhouette techniques. This reconstruction technique produces objects with small synthetic bulges, making them more difficult to align accurately. On the one hand, prealignment and branch-and-bound techniques are used to improve the convergence and speed of the alignment algorithms. On the other hand, a method to obtain a precise alignment even in the presence of bulges is presented. Finally, a refinement of the shape-from-silhouettes technique is shown. This technique uses multiple captures to refine object reconstruction and reduce or eliminate, among other improvements, synthetic bulges.

Industrial inspection is a well-known application area for unmanned aerial vehicles (UAVs), but are modern market-available drones fully suitable for inspections of larger-scale industrial facilities? This review summarizes the pros and cons of aerial large-scale facility inspection, distinguishing it from other inspection scenarios implemented with drones. Moreover, based on paper analysis and additionally performed experimental studies, it reveals specific issues related to modern commercial drone software and demonstrates that market-available UAVs (including DJI and Autel Robotics) more or less suffer from the same problems. The discovered issues include a Global Navigation Satellite System (GNSS) Real Time Kinematic (RTK) shift, an identification of multiple images captured from the same point, limitations of custom mission generation with external tools and mission length, an incorrect flight time prediction, an unpredictable time of reaching a waypoint with a small radius, deviation from the pre-planned route line between two waypoints, a high pitch angle during acceleration/deceleration, an automatic landing cancellation in a strong wind, and flight monitoring issues related to ground station software. Finally, on the basis of the paper review, we propose solutions to these issues, which helped us overcome them during the first autonomous inspection of a 2400 megawatts thermal power plant.

Anomaly detection with machine learning in industrial inspection systems for manufactured products relies on labelled data. This raises the question of how the labelling by humans should be conducted. Moreover, such a system will most likely always be imperfect and potentially need a human fall-back mechanism for ambiguous cases. We consider the case where we want to optimise the cost of the combined inspection process done by humans together with a pre-trained algorithm. This gives improved combined performance and increases the knowledge of the performance of the pre-trained model. We focus on so-called one-class classification problems which produce a continuous outlier score. After establishing some initial setup mechanisms ranging from using prior knowledge to calibrated models, we then define some cost model for machine inspection with a possible second inspection of the sample done by a human. Further, we discuss in this cost model how to select two optimal boundaries of the outlier score, where in between these two boundaries human inspection takes place. Finally, we frame this established knowledge into an applicable algorithm and conduct some experiments for the validity of the model. [ABSTRACT FROM AUTHOR]

We propose a novel deep learning-based anomaly detection (AD) system that combines a pixelwise classification network with conditional normalizing flow (CNF) networks by sharing feature extractors. We trained the pixelwise classification network using synthetic abnormal data to fine-tune a pretrained feature extractor of the CNF networks, thereby learning the discriminative features of the in-domain data. After that, we trained the CNF networks using normal data with the fine-tuned feature extractor to estimate the density of normal data. During inference, we detected anomalies by calculating the weighted average of the anomaly scores from the pixelwise classification and CNF networks. Because the proposed system not only has learned the properties of in-domain data but also aggregated the anomaly scores of the classification and CNF networks, it showed significantly improved performance compared to existing methods in experiments using the MvTecAD and BTAD datasets. Moreover, the proposed system does not increase computations intensively since the classification and the density estimation systems share feature extractors.

Diagnosis of complex engineering systems requires the use of multiple sensor sources to acquire information. In this paper we present a survey of multimodal data acquisition systems for nondestructive testing (NDT) and engineering analysis. We begin with a summary of the relative strengths and weaknesses of individual NDT modalities. Thereafter we present existing multimodal inspection hardware systems that use complementary NDT sensors. The advantages of such multimodal data acquisition over conventional single modality sensors in inspection and analysis are highlighted. Possible approaches to fuse complementary multimodal sensor data are discussed. We conclude with possible directions for the future development of multimodal inspection systems. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This work addresses the problem of semi-automatic inspection and navigation in confined environments. A system that overcomes many challenges at the state of the art is presented. It comprises a multirotor able to inspect an industrial combustion chamber thus working in a GPS-denied environment with poor lighting conditions, in the presence of magnetic and communication disturbances, iron dust and repetitive patterns on the structure walls. The presented system is able to pass through narrow entrances but still capable of acquiring high resolution images and to allow operators to perform inspection of the structures. Starting from the captured data, the system is able to provide a 3D reconstruction of the inspected environment for offline analysis. [ABSTRACT FROM AUTHOR]