Wyniki wyszukiwania - BazTech

1

Kompresja wizji dla maszyn

Domański Marek

Przegląd Telekomunikacyjny + Wiadomości Telekomunikacyjne

|

2025

|

nr 4

32--36

PL

Metody kompresji wizji są opracowywane tak, by przy zadanej małej bitowej szybkości transmisji dawało się w odbiornikach uzyskiwać możliwie dobrą jakość zdekodowanych obrazów. Dotychczas miarą tej jakości była uśredniona ocena grupy widzów. Jednak w ostatniej dekadzie coraz częstsze są zastosowania, w których obraz jest poddawany kompresji przed wykonaniem różnych operacji widzenia maszynowego. Powstaje więc pytanie, czy dotychczasowe metody kompresji zoptymalizowane pod względem subiektywnie ocenianej przez ludzi jakości zdekodowanych obrazów są także optymalne dla kompresji wizji przeznaczanej do wykorzystania w systemach widzenia maszynowego. Odpowiedź na to pytanie okazuje się być negatywna i dlatego w ostatnim dziesięcioleciu prowadzi się intensywne badania nad nowymi metodami kompresji odpowiednimi dla wizji używanej przez systemy widzenia maszynowego, czyli nad metodami kompresji wizji dla maszyn (video compression/coding for machines). Praca skrótowo opisuje rezultaty badań dotyczących metod kompresji wizji dla maszyn. W szczególności podano najważniejsze cechy kodeka MPEG MVC, który został ujęty w projekcie części drugiej normy ISO/IEC MPEG-AI.

EN

Until recently, video coding standards were developed from the perspective of a human observer; that is, the main goal was to achieve a high subjective quality of the decoded video. This quality was measured by mean opinion score. However, over the past decade, the development of semi- and fully autonomous vehicles, AI-based video analysis, intelligent video surveillance, and video-based control in many new areas has led to a sharp increase in the amount of video data shared between computers. In these cases, direct human consumption of the decoded video is not the primary application. These factors have sparked research interest in video coding where the decoded data serves as input for machine vision tasks. This paper briefly summarizes the results of research on video coding for machines. Specifically, it briefly describes the upcoming MPEG standard for machine video coding as part two of the ISO/IEC MPEG-AI standard.

2

A Machine Vision System for Inspecting Mechanical Parts

Rajagounder Rajamani

Machine Graphics and Vision

|

2025

|

Vol. 34, No. 1

75--86

EN

Computer vision-based inspection has become widely used in manufacturing industries for part identification, dimensional inspection, and guiding material handling systems. Defect-free production cannot be achieved with sampling inspection methods; therefore, a 100 percentage inspection approach is mandatory to meet the zero-defect goals of manufacturing industries. Achieving this is possible with advanced technologies, such as vision-based inspection systems. In this study, a vision-based inspection system is proposed for part identification, defect detection, and dimensional measurement. The system is validated using machined parts, including a Druck plate, Pressure plate, and Retainer. A part identification algorithm is developed based on a geometry search approach. The inspection algorithm classifies parts based on edge relationships, utilizing edge detection techniques to identify each part’s geometric features. Surface defects are identified by analyzing the pixel intensity gradients within defective regions. The system measures part dimensions using a vision system, with results comparable to those obtained from a coordinate measuring machine.

3

Machine vision-based surface defect detection method for welds

Xu Su, Cui Wenxuan, Zhou Xianzhang, Zhong Qiucheng, Wei Yifan, Wang Yipeng

Diagnostyka

|

2025

|

Vol. 26, No. 1

art. no. 2025106

EN

The texture feature extraction including grayscale co-occurrence matrix and various shape feature extraction methods are adopted in this paper, as well as convolutional neural network based on Visual Geometry Group-16 structure. In particular, the Squeeze-and-Excitation module and dilated convolution technique are introduced to improve the model, aiming to enhance its feature extraction and classification capabilities. On the JPEGWELD dataset, the improved model had 98.7% accuracy in the training set, 97.9% accuracy in the test set, and 98.7% recall rate. In the comparative analysis, although the number of parameters of the improved VGG16 model was 33.64M and the maximum model size was 385MB, the detection time was only 1.3s. The results demonstrated that the model had efficient optimization and computational performance, with a good balance between design and optimization while maintaining a short detection time. The proposed method exhibits high accuracy and efficiency in the detection of various types of weld defects, demonstrating strong universality and adaptability. Its applicability to diverse industrial settings is evident. The study provides an effective solution for industrial automated inspection, which is of great significance to improve the quality control level and production efficiency of manufacturing industry.

4

Design of automatic target recognition and detection system for sensors based on improved machine vision control

Zhou Wei, Xu Ye, Huang Yue

Diagnostyka

|

2025

|

Vol. 26, No. 2

art. no. 2025206

EN

The aim of this study is to design and implement a sensor-based automatic target recognition and detection system with machine vision control. The system achieves high-precision detection of targets in complex environments by integrating multiple sensors, including industrial-grade color and infrared cameras, VelodyneHDL-64E lidar, ultrasonic arrays, and high-quality IMU devices. Through multi-sensor data fusion, preprocessing techniques, and feature extraction methods, the experimental results show that the system is able to achieve high-precision target detection in different scenarios. FasterR-CNN and its improved version of the model perform well in the experiments, especially after the introduction of the feature pyramid network (FPN) and the attention mechanism, which significantly improves the detection rate and the overall performance of the small targets. Experimental results show that the multi-sensor fusion system significantly improves the performance in target detection, with the accuracy of RGB cameras increasing from 85% to 92% and the recall rate increasing from 78% to 88%. After introducing the feature pyramid network (FPN) and attention mechanism, the detection accuracy of the Faster R-CNN model for small targets increased from 70% to 75%. Although the processing speed decreased slightly (from 20fps to 15fps), the overall detection accuracy and robustness were significantly enhanced. In addition, the model pruning technology increased the processing speed to 12fps while maintaining high accuracy, which is suitable for real-time applications. The model pruning technique successfully realizes the lightweighting of the model while maintaining high detection accuracy, which provides the possibility of real-time target detection for embedded devices.

5

Convolutional neural networks for froth imaging in a flotation machine with electromagnetic regrinding impeller

Ogonowski Szymon, Wójtowicz Paweł

Archives of Control Sciences

|

2025

|

Vol. 35, No. 4

771--792

EN

This study investigates a vision-based supervisory framework for a prototype flotation machine equipped with an in-line electromagnetic regrinding impeller, a configuration intended to enhance mineral liberation and recovery for fine particles and tailings. Building on recent advances in machine vision for process industries, convolutional neural networks (CNNs) were trained to infer supervisory variables directly from froth images: (i) bubble population metrics (bubble count/density) and (ii) overall concentrate yield. An image corpus of >1000 froth frames was collected across three trials under deliberately varied operating conditions (airflow, electromagnetic mill frequency, and feed mass), yielding substantial covariate shift. The results provide preliminary evidence that CNN-based froth imaging can supply actionable supervisory signals in electromagnetic-assisted flotation. Future work will expand datasets, harden illumination robustness, incorporate spatiotemporal modeling and sensor fusion, and evaluate closed-loop control in prospective trials.

6

Integrating YOLOv5, Jetson nano microprocessor, and Mitsubishi robot manipulator for real-time machine vision application in manufacturing: A lab experimental study

Kirda Ardian Webi, Majewski Paweł, Bursy Gerard, Bartoszuk Marian, Yassin Hayati, Królczyk Grzegorz, Akbar Nur Arifin, Caesarendra Wahyu

Advances in Science and Technology. Research Journal

|

2025

|

Vol. 19, no. 5

248--270

EN

Efficient detection and rectification of metal components conditions during manufacturing and post-processing manufacturing are crucial for quality control in industries. This paper describes a lab-scale integrated system for real-time and auto-mated metal edge image detection using YOLOv5 machine vision algorithm for automated met-al grinding and chamfering in manufacturing. The YOLOv5 algorithm was compared with VGG-16 and ResNet algorithm for edge detection i.e., sharp edge, chamfer edge, and burrs edge on the metal workpiece. The YOLOv5 algorithm and model were developed and embedded in the NVIDIA Jetson Nano microprocessor. An integrated system connects the NVIDIA Jetson Nano microprocessor with an embedded deep learning image processing model to a Mitsubishi Electric Melfa RV-2F-1D1-S15 robot manipulator to perform the lab-scale manufacturing process for automated grinding and chamfering. The models demonstrates durable performance in detecting the metal edge image for intelligent manufacturing application, achieving a mean average precision 0.854 for ResNet, 0.942 for VGG-16 and 0.957 for YOLOv5, all models across defect classes with minimal misclassifications. The Mitsubishi Electric Melfa RV-2F-1D1-S15 robot manipulator received input from the machine vision system and per-formed an automated grinding and chamfering process accordingly; By integrating camera, embedded deep learning in the microprocessor and robot manipulator, auto-mated grinding and chamfering process in metal edge component can be efficiently rectified. This machine vision technology tailored solution promises to improve productivity and consistency in metal component manufacturing.

7

Application of YOLOv8 in fusion welding defect detection on carbon steel for potential remote visual inspection

Rappe Nurham Ambo, Kirda Ardian Webi, Yassin Hayati, Bartoszuk Marian, Caesarendra Wahyu

Advances in Science and Technology. Research Journal

|

2025

|

Vol. 19, no. 12

436--453

EN

Deep learning and machine vision technologies nowadays is a power artificial intelligence-based on industrial welding defect detection operations in manufacturing and require superior automated inspection systems. YOLOv8 represents the advanced stage of YOLO deep neural network architecture which brings powerful object detection features to fusion welding applications resulting in high accuracy for computer vision quality control technology. The integration between artificial intelligence and traditional inspection approaches now provides a viable route for reducing dependency on humans through automated methods that inspect conventional welds. The research executes YOLOv8 as a leading-edge deep learning structure which detects welding flaws automatically through machine vision systems for potential remote welding surface assessments. The proposed system stands apart from previously described systems because it combines high-resolution machine vision cameras with the YOLOv8 sophisticated convolutional neural network structure. Standardised remote visual inspection configurations were implemented to gather datasets from steel weld inspections while testing the system for various typical carbon steel welding defect shapes. The training portion of the deep learning model underwent evaluations in detail to assess both its real time deployment suitability and its defect identification and categorisation abilities. The testing process verified remarkable system performance with high accuracy reaching 98% confidence for its complex deep learning algorithms. The system offers better assessment speed than traditional inspection methods and simultaneously lowers the need for human involvement and offers total digital documentation for quality control purposes. The results demonstrate YOLOv8 to be a potential leading technology for the next-generation of industrial welding quality control systems as it persists robust surface defect identification across multiple inspection conditions.

8

Research on light weight intelligent identification method of coal and gangue

Li Deyong, Wang Guofa, Guo Yongcun, Yang Yuhao, Wang Shuang, Wang Wenshan

Physicochemical Problems of Mineral Processing

|

2024

|

Vol. 60, iss. 2

art. no. 187975

EN

Accurate identification of coal and gangue is essential for clean and efficient use of coal. Existing target detection algorithms are ineffective in detecting small-target and overlapping gangue, and contain complex network structure and large parameter volume, which cannot meet the demand of real-time detection of edge devices. To address the above problems, a lightweight detection and identification approach of coal gangue based on improved YOLOv5s is proposed. The depth-separable convolutions are used to replace ordinary convolutions, and the C3 (Concentrated-Comprehensive Convolution Block) Ghost module is constructed to replace all C3 modules in the YOLOv5s to reduce model computation and parameters. The CA (Coordinate Attention) attention mechanism is introduced to strengthen the attention to the target to be detected, suppress irrelevant background interference, and improve the detection accuracy of the model. The Focal- EIOU (Focal and Efficient Intersection Over Union) loss function was introduced to replace the original CIOU. Extensive experiments substantiated the proposed approach can effectively and quickly detect the small-target and overlapping coal gangue accurately, and the mAP (mean Average Presicion) reaches 97.7%. Compared with the original YOLOv5s, the proposed approach reduces the number of parameters and the amount of computation by 48.5% and 43%, respectively, under the premise of maintaining the same detection accuracy.

9

AI functionalities in cobot-based manufacturing for performance improvement in quality control application

Moor Madis, Sarkans Martins, Kangru Tavo, Otto Tauno, Riives Juri

Journal of Machine Engineering

|

2024

|

Vol. 24, No. 2

44-55

EN

Modern manufacturing faces vastly changing challenges. The current economic situation and technological developments in terms of Industry 4.0 (I4.0) and Industry 5.0 (I5.0) force enterprises to integrate new technologies for more efficient and higher-quality products. Artificial intelligence (AI) and Machine Learning (ML) are the technologies that make machines capable of making human-like decisions. In the long run, AI and ML can add a layer (functionality) to make IoT devices more interactive and user-friendly. These technologies are driven by data and ML uses different types of data for making decisions. Our research focuses on testing a cobot-based quality control (CBQC) system that uses smart fixture and machine vision (MV) to determine the cables inside products with similar designs, but different functionality. The products are IoT modules for small electric vehicles used for interface, connectivity, and GPS monitoring. Previous research describes the methodology of reconfiguration of existing cobot cells for quality control purposes. In this paper, we discuss the testing of the CBQC system, together with creating a pattern database, training the ML model, and adding a predictive model to avoid defects in product cable sequence. Preliminary testing is carried out in the laboratory environment which leads to production testing in SME manufacturing. Results, developments, and future work will be presented at the end of the paper.

10

Detection of three-dimensional belt deviation and longitudinal tearing defects based on binocular line laser technology

Zhang Zhenming, Yang Changjun, Huang Chunhua, Gao Hong, Bao Yongtao

Diagnostyka

|

2024

|

Vol. 25, No. 4

art. no. 2024401

EN

The detection of belt deviation and longitudinal tearing defects is the key to ensuring the safe and reliable operation of the equipment. A three-dimensional belt deviation and longitudinal tear defect detection system based on binocular line laser technology was proposed to address the low detection efficiency and high delay of conveyor belt deviation and longitudinal tear detection. A line laser was irradiated onto the surface of the belt through an image acquisition device, and the collected images were preprocessed. Image segmentation, feature extraction, and pattern recognition techniques were used to detect belt deviation and longitudinal tearing defects. These results confirmed that the system designed in this study only took an average of 20 to 30 milliseconds to process an image. The average accuracy of secondary detection was 97.37%, which was 7.5% higher than that of primary detection. The average processing time of the first level detection was 19.45ms. The average processing time of the two level detection was 23.73ms, which was 4.28ms longer than the first level detection. The designed 3D belt deviation and longitudinal tearing defect detection system based on binocular laser technology has high real-time and accuracy, which is very important for the safety production of enterprises.

11

SmokeFinder - wspomagany algorytmami sztucznej inteligencji system wizyjnej detekcji dymu

Zuchniak Konrad, Kosk Eryk, Jaglarz Piotr

Przegląd Telekomunikacyjny + Wiadomości Telekomunikacyjne

|

2023

|

nr 4

194--197

PL

SmokeFinder to system umożliwiający stałe analizowanie obrazu pochodzącego z kamer obserwacyjnych w celu wyszukiwania dymu w lesie. W przypadku wykrycia nawet niewielkich słupków dymu do operatora systemu wysyłane są ostrzeżenia. Zaawansowane algorytmy uczenia maszynowego analizują dane zebrane z kamery obserwacyjnej w czasie rzeczywistym. W przypadku wykrycia dymu ustalana jest jego lokalizacja. Dane z wielu kamer są łączone w celu usprawnienia akcji gaśniczej. W przypadku potwierdzenia zagrożenia automatycznie zaalarmowane zostaną odpowiednie lokalne jednostki straży pożarnej.

EN

SmokeFinder is a system that allows to constantly analyze the image from surveillance cameras in order to search for smoke in the forest. If even small pillars of smoke are detected, alerts are sent to the system operator. Advanced machine learning algorithms analyze the data collected from the observation camera in real time. If smoke is detected, its location is determined. Data from multiple cameras are combined to improve firefighting. If the threat is confirmed, the relevant local fire departments will be automatically alerted.

12

Features of the implementation of computer vision in the problems of automated product quality contro

Stelmakh Nataliia, Mastenko Ihor, Sulima Olga, Rudyk Tetiana

Informatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska

|

2023

|

T. 13, nr 1

38--41

EN

The article analyzes the fields of application of machine vision. Special attention is focused on the application of Machine Vision in intelligent technological systems for product quality control. An important aspect is a quick and effective analysis of product quality directly at the stage of the technological process with high accuracy in determining product defects. The appropriateness and perspective of using the mathematical apparatus of artificial neural networks for the development of an intelligent technological system for monitoring the geometric state of products have been demonstrated. The purpose of this study is focused on the identification and classification of reed tuber quality parameters. For this purpose, new methods of identification and classification of quality control of various types of defects using computer vision and machine learning algorithms were proposed.

PL

W artykule dokonano analizy obszarów zastosowań widzenia maszynowego. Szczególną uwagę zwrócono na zastosowanie widzenia maszynowego w inteligentnych systemach technologicznych kontroli jakości wyrobów. Ważnym aspektem jest szybka i skuteczna analiza jakości produktu bezpośrednio na etapie procesu technologicznego z dużą dokładnością w określaniu wad produktu. Pokazano celowość i perspektywę wykorzystania aparatu matematycznego sztucznych sieci neuronowych do budowy inteligentnego systemu technologicznego do monitorowania stanu geometrycznego wyrobów. Celem badań jest identyfikacja i klasyfikacja parametrów jakościowych rurek trzcinowych. W tym celu zaproponowano nowe metody identyfikacji i klasyfikacji kontroli jakości różnego rodzaju defektów z wykorzystaniem wizji komputerowej i algorytmów uczenia maszynowego.

13

Machine vision online detection of ore granularity based on edge computing

Yao Jiang, Xue Yinbo, Li Xiaoliang, Zhai Lei, Yang Zhenyu, Zhang Wenhui

Archives of Mining Sciences

|

2023

|

Vol. 68, no. 2

335--350

EN

Belts are widely applied in mine production for conveying ores. Understanding ore granularity, which is a crucial factor in determining the effectiveness of crushers, is vital for optimising production efficiency throughout the crushing process and ensuring the success of subsequent operations. Based on edge computing technology, an online detection method is investigated to rapidly and accurately obtain ore granularity information on high-speed conveyor belts. The detection system utilising machine vision technology is designed in this paper. The high-speed camera set above the belt is used to collect the image of the ore flow, and the collected image is input into the edge computing device. After binary, grey morphology and convex hull algorithm processing, the particle size distribution of ore is obtained by statistical analysis. Finally, a 5G router is used to output the settlement result to a cloud platform. In the GUANBAOSHAN mine of Ansteel Group, the deviation between manual screening and image particle size analysis was studied. Experimental results show that the proposed method can detect the ore granularity, ore flow width and ore flow terminal in real-time. It can provide a reference for the staff to adjust the parameters of the crushing equipment, reduce the mechanical loss and the energy consumption of the equipment, improve the efficiency of crushing operation and reduce the failure rate of the crusher.

14

Industrial Application of Surface Crack Detection in Sheet Metal Stamping Using Shift-and-Add Speckle Imaging

Świłło Sławomir, Cacko Robert

Advances in Science and Technology. Research Journal

|

2023

|

Vol. 17, no. 5

220--235

EN

The sheet metal surface crack detection during manufacturing is an essential issue because of both the product quality and process productivity. Development of solutions to eliminate defective products during the metal forming process is crucial for the smooth production and for developing an appropriate tool geometry in the initial phase of the process. Currently, the methods of surface crack detection used in the industry are mostly related to visual inspection. These are methods that require operators of industrial facilities considerable attention and effort to capture emerging discontinuities on the sheet metal surface. Also, this situation results increase in the duration of the specific operations of stamping and significantly reduces productivity. Therefore, an industrial application of a non-contact laser technique that simultaneously provides the results of the speckle imaging is presented. The authors demonstrate a specially designed machine vision system along with experimental tools for the stamping operation. Proposed solution uses the phenomenon of speckle pattern that appears in the image of the investigated sheet surface produced by the laser beam emission. In this method, coherent laser light is emitted to the surface, where a speckle pattern is generated due to scatter reflection from the sheet metal surface and then, shift-and-add technique and image processing is applied. The proposed measurement technique consists, initially, of making a sequence of images of the tested object for the moving surface of the sheet. Secondly, the object's displacement quantity in each image is determined, and the position is corrected. The test object in each image is moved to the starting position, and all images are superimposed. It allows to obtain a high-quality image with visible surface defects. Finally, the dynamically changing speckle pattern intensity is evaluated using Gaussian-of-Laplacian edge detection to investigate a surface crack location due to the surface discontinues and light scattering. This process is recommended for machine vision imaging of distant objects, which works well in industrial conditions as well as online analysis. Also, from the speckle size measurement, an experimental procedure is employed to verify the best condition for vision system resolution.

15

Enhanced of Tool Management Based on Machine Vision in the Field of Metal Forming Technology

Świłło Sławomir, Cacko Robert

Archives of Metallurgy and Materials

|

2022

|

Vol. 67, iss. 4

1419--1425

EN

In this article, the authors focused on the widely used aluminium extrusion technology, where the die quality and durability are the essential factors. In this study, detailed solutions in the three-key area have been presented. First is applying marking technology, where a laser technique was proposed as a consistent light source of high power in a selected, narrow spectral range. In the second, an automated and reliable identification method of alphanumeric characters was investigated using an advanced machine vision system and digital image processing adopted to the industrial conditions. Third, a proposed concept of online tool management was introduced as an efficient process for properly planning the production process, cost estimation and risk assessment. In this research, the authors pay attention to the designed vision system’s speed, reliability, and mobility. This leads to the practical, industrial application of the proposed solutions, where the influence of external factors is not negligible.

16

New approach to spindle thermal extension measuring based on machine vision for the vertical maching centre

Su Dongxu, Cai Xin, Li Yang, Zhao Wanhuan, Zhang Huijie

Metrology and Measurement Systems

|

2021

|

Vol. 28, nr 2

357--370

EN

When machine tool spindles are running at a high rotation speed, thermal deformation will be introduced due to the generation of large amounts of heat, and machining accuracy will be influenced as a result, which is a generalized issue in numerous industries. In this paper, a new approach based on machine vision is presented for measurements of spindle thermal error. The measuring system is composed of a Complementary Metal-Oxide-Semiconductor (CMOS) camera, a backlight source and a PC. Images are captured at different rotation angles during end milling process. Meanwhile, the Canny edge detection and Gaussian sub-pixel fitting methods are applied to obtain the bottom edge of the end mill which is then used to calculate the lowest point coordinate of the tool. Finally, thermal extension of the spindle is obtained according to the change of the lowest point at different time steps of the machining process. This method is validated through comparison with experimental results from capacitive displacement sensors. Moreover, spindle thermal extension during the processing can be precisely measured and used for compensation in order to improve machining accuracy through the proposed method.

17

An approach to an intelligent scanning of the machine tool workspace

Marchewka Lukasz, Grudzinski Marek

Journal of Machine Engineering

|

2021

|

Vol. 21, No. 2

60--74

EN

Modern 3D scanners can measure the geometry with high accuracy and within a short time. In turn, currently produced CNC machine tools allow for very accurate manufacturing; however, processes beyond the machining cycle remain time-consuming. This paper presents the idea and experimental tests of the scanning system in the CNC machine, which allows to speed up on-machine measurements, align clouds of 3D data points with an accuracy close to that of the machine itself, and finally set the workpiece coordinate system for machining. This modern approach is in line with Industry 4.0, combining the terms of data processing, machine vision, manufacturing automation, and human-machine interfaces. The future implementation of the proposed system as an interchangeable tool will allow performing autonomous measurements, inspection, and supervision of the workspace, without engaging the machine operator. The system calibration and experimental results using the industrial 3D scanner and CNC machine are described.

18

A new approach to the histogram-based segmentation of synthetic aperture radar images

Siemiatkowska Barbara, Gromada Krzysztof

Journal of Automation Mobile Robotics and Intelligent Systems

|

2021

|

Vol. 15, No. 1

39--42

EN

Radar machine vision is an emerging research field in the mobile robotics. Because Synthetic Aperture Radars (SAR) are robust against weather and light condition, they pro‐ vide more useful and reliable information than optical images. On the other hand, the data processing is more complicated and less researched than visible light images processing. The main goal of our research is to build sim‐ ple and efficient method of SAR image analysis. In this ar‐ ticle we describe our research related to SAR image seg‐ mentation and attempts to detect elements such as the buildings, roads and forest areas. Tests were carried out for the images made available by Leonardo Airborne & Space System Company.

19

Machine vision in autonomous vehicles: designing and testing the decision making algorithm based on entity attribute value model

Shubenkova Ksenia, Zabinski Krzysztof

Archiwum Motoryzacji

|

2021

|

Vol. 94, nr 4

27--37

EN

If we speak about the Smart City’s transport system, autonomous vehicles idea is the first thing that comes to mind. Today, it is strongly believed that the autonomous vehicles’ introduction into the traffic will increase the road safety. However, driverless cars are not the solution by itself. The road safety and, accordingly, sustainability will strongly depend on decision making algorithms inbuilt into the control module. Therefore, the goal of our research is to design and test the data mining algorithm based on Entity–Attribute–Value (EAV) model for decision making in the Intelligent System in the fully- or semi-autonomous vehicles. In this article, we describe the methodology to create 3 main modules of the designed Intelligent System: (1) an Object detection module; (2) a Data analysis module; (3) a Knowledge database built on decision rules generated with the help of our data mining algorithm. To build the Decision Table on the base of the real data, we have tested our algorithm on a simple collection of photos from a Polish two-lane road. Generated rules provide comparable classification results to the dynamic programming approach for optimization of decision rules relative to length or support. However, our decision making algorithm thanks to excluding the mistakes made on the object detection stage, works faster than existing ones with the same level of correctness.

20

Positioning error correction of autonomously movable robot arm

Vaher Kristo, Otto Tauno, Riives Jüri

Journal of Machine Engineering

|

2020

|

Vol. 20, No. 4

152--160

EN

Industrial robots are mainly used stationarily in one working position. SMEs often find themselves in situations where robots don’t have enough work to do, and because in general, robots cannot be easily moved to another position, the efficiency of robots will decrease. This study provides a solution for this issue. The solution can be found in a robot work cell where a mobile robot deals with robot arm transportation. However, since the mobile robot is not precise enough in positioning, machine vision is used to overcome this problem, which helps the robot to position itself accurately in relation to the work object. The solution has been developed and tested successfully at an Industry 4.0 testbed.