Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm
Objects tracking system is a research field of great importance. The two main reasons for this: 1 - Attempting to replace humans with robots or automatic systems in every field wherever possible. 2 - The accuracy, effectiveness and speed of those systems compared to human performance. In this resear...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
FRUCT
2020-09-01
|
| Series: | Proceedings of the XXth Conference of Open Innovations Association FRUCT |
| Subjects: | |
| Online Access: | https://www.fruct.org/publications/fruct27/files/Sha.pdf |
| _version_ | 1818452254781341696 |
|---|---|
| author | Ali Shakkouf Vladislav Gromov |
| author_facet | Ali Shakkouf Vladislav Gromov |
| author_sort | Ali Shakkouf |
| collection | DOAJ |
| description | Objects tracking system is a research field of great importance. The two main reasons for this: 1 - Attempting to replace humans with robots or automatic systems in every field wherever possible. 2 - The accuracy, effectiveness and speed of those systems compared to human performance. In this research we propose a new approach for tracking objects and predicting the future trajectory of these objects. the main idea is to observe the object for about 15 seconds, this gives us part of the moving pattern, then Dynamic Regressor Extension and Mixing (DREM) algorithm is used to estimate the main frequencies involved in the detected movement pattern. The estimated frequencies are used to predict the future trajectory of the object. The accuracy of this method was tested using 5DOF arm robot, which try to grip the object at each moment of its future trajectory. This research performed on a virtual object designed in vredit in MATLAB. The object moves on an LCD screen. The results were represented as the difference between the predicted trajectory and the real future trajectory. Results show that tracking error of a ball moves on the screen for different moving pattern is less than 1 cm, where the screen stands 500 mm far from arm base. |
| first_indexed | 2024-12-14T21:20:10Z |
| format | Article |
| id | doaj.art-127ce888cf8c4446afd460503ff7fcdc |
| institution | Directory Open Access Journal |
| issn | 2305-7254 2343-0737 |
| language | English |
| last_indexed | 2024-12-14T21:20:10Z |
| publishDate | 2020-09-01 |
| publisher | FRUCT |
| record_format | Article |
| series | Proceedings of the XXth Conference of Open Innovations Association FRUCT |
| spelling | doaj.art-127ce888cf8c4446afd460503ff7fcdc2022-12-21T22:46:58ZengFRUCTProceedings of the XXth Conference of Open Innovations Association FRUCT2305-72542343-07372020-09-0127121422010.23919/FRUCT49677.2020.9211083Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM AlgorithmAli Shakkouf0Vladislav Gromov1ITMO University, RussiaITMO University, RussiaObjects tracking system is a research field of great importance. The two main reasons for this: 1 - Attempting to replace humans with robots or automatic systems in every field wherever possible. 2 - The accuracy, effectiveness and speed of those systems compared to human performance. In this research we propose a new approach for tracking objects and predicting the future trajectory of these objects. the main idea is to observe the object for about 15 seconds, this gives us part of the moving pattern, then Dynamic Regressor Extension and Mixing (DREM) algorithm is used to estimate the main frequencies involved in the detected movement pattern. The estimated frequencies are used to predict the future trajectory of the object. The accuracy of this method was tested using 5DOF arm robot, which try to grip the object at each moment of its future trajectory. This research performed on a virtual object designed in vredit in MATLAB. The object moves on an LCD screen. The results were represented as the difference between the predicted trajectory and the real future trajectory. Results show that tracking error of a ball moves on the screen for different moving pattern is less than 1 cm, where the screen stands 500 mm far from arm base.https://www.fruct.org/publications/fruct27/files/Sha.pdfcomputer vision systemmoving objects trackingfrequency estimationdrem algorithm |
| spellingShingle | Ali Shakkouf Vladislav Gromov Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm Proceedings of the XXth Conference of Open Innovations Association FRUCT computer vision system moving objects tracking frequency estimation drem algorithm |
| title | Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm |
| title_full | Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm |
| title_fullStr | Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm |
| title_full_unstemmed | Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm |
| title_short | Computer Vision System to Track Moving Objects With unknown Periodic Moving Patterns Based on DREM Algorithm |
| title_sort | computer vision system to track moving objects with unknown periodic moving patterns based on drem algorithm |
| topic | computer vision system moving objects tracking frequency estimation drem algorithm |
| url | https://www.fruct.org/publications/fruct27/files/Sha.pdf |
| work_keys_str_mv | AT alishakkouf computervisionsystemtotrackmovingobjectswithunknownperiodicmovingpatternsbasedondremalgorithm AT vladislavgromov computervisionsystemtotrackmovingobjectswithunknownperiodicmovingpatternsbasedondremalgorithm |