Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures
The emerging satellite videos provide the opportunity to detect moving objects and track their trajectories, which were not possible for remotely sensed imagery with limited temporal resolution. So far, most studies using satellite video data have been concentrated on traffic monitoring through dete...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-11-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/12/21/3589 |
| _version_ | 1797549111288791040 |
|---|---|
| author | Fan Shi Fang Qiu Xiao Li Ruofei Zhong Cankun Yang Yunwei Tang |
| author_facet | Fan Shi Fang Qiu Xiao Li Ruofei Zhong Cankun Yang Yunwei Tang |
| author_sort | Fan Shi |
| collection | DOAJ |
| description | The emerging satellite videos provide the opportunity to detect moving objects and track their trajectories, which were not possible for remotely sensed imagery with limited temporal resolution. So far, most studies using satellite video data have been concentrated on traffic monitoring through detecting and tracking moving cars, whereas the studies on other moving objects such as airplanes are limited. In this paper, an integrated method for monitoring moving airplanes from a satellite video is proposed. First, we design a normalized frame difference labeling (NFDL) algorithm to detect moving airplanes, which adopts a non-recursive strategy to deliver stable detection throughout the whole video. Second, the template matching (TM) technique is utilized for tracking the detected moving airplanes in the frame sequence by improved similarity measures (ISMs) with better rotation invariance and model drift suppression ability. Template matching with improved similarity measures (TM-ISMs) is further implemented to handle the leave-the-scene problem. The developed method is tested on a satellite video to detect and track eleven moving airplanes. Our NFDL algorithm successfully detects all the moving airplanes with the highest F<sub>1</sub> score of 0.88 among existing algorithms. The performance of TM-ISMs is compared with both its traditional counterparts and other state-of-the-art tracking algorithms. The experimental results show that TM-ISMs can handle both rotation and leave-the-scene problems. Moreover, TM-ISMs achieve a very high tracking accuracy of 0.921 and the highest tracking speed of 470.62 frames per second. |
| first_indexed | 2024-03-10T15:09:07Z |
| format | Article |
| id | doaj.art-79e5a306fed9491dad550fe46d670e97 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T15:09:07Z |
| publishDate | 2020-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-79e5a306fed9491dad550fe46d670e972023-11-20T19:27:48ZengMDPI AGRemote Sensing2072-42922020-11-011221358910.3390/rs12213589Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity MeasuresFan Shi0Fang Qiu1Xiao Li2Ruofei Zhong3Cankun Yang4Yunwei Tang5Geospatial Information Sciences, The University of Texas at Dallas, 800 West Campbell Road, Richardson, TX 75080, USAGeospatial Information Sciences, The University of Texas at Dallas, 800 West Campbell Road, Richardson, TX 75080, USAGeospatial Information Sciences, The University of Texas at Dallas, 800 West Campbell Road, Richardson, TX 75080, USABeijing Advanced Innovation Center for Imaging Technology, Capital Normal University, Beijing 100048, ChinaBeijing Advanced Innovation Center for Imaging Technology, Capital Normal University, Beijing 100048, ChinaKey Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaThe emerging satellite videos provide the opportunity to detect moving objects and track their trajectories, which were not possible for remotely sensed imagery with limited temporal resolution. So far, most studies using satellite video data have been concentrated on traffic monitoring through detecting and tracking moving cars, whereas the studies on other moving objects such as airplanes are limited. In this paper, an integrated method for monitoring moving airplanes from a satellite video is proposed. First, we design a normalized frame difference labeling (NFDL) algorithm to detect moving airplanes, which adopts a non-recursive strategy to deliver stable detection throughout the whole video. Second, the template matching (TM) technique is utilized for tracking the detected moving airplanes in the frame sequence by improved similarity measures (ISMs) with better rotation invariance and model drift suppression ability. Template matching with improved similarity measures (TM-ISMs) is further implemented to handle the leave-the-scene problem. The developed method is tested on a satellite video to detect and track eleven moving airplanes. Our NFDL algorithm successfully detects all the moving airplanes with the highest F<sub>1</sub> score of 0.88 among existing algorithms. The performance of TM-ISMs is compared with both its traditional counterparts and other state-of-the-art tracking algorithms. The experimental results show that TM-ISMs can handle both rotation and leave-the-scene problems. Moreover, TM-ISMs achieve a very high tracking accuracy of 0.921 and the highest tracking speed of 470.62 frames per second.https://www.mdpi.com/2072-4292/12/21/3589satellite videosmoving airplane detectionmoving airplane trackingtemplate matchingsimilarity measures |
| spellingShingle | Fan Shi Fang Qiu Xiao Li Ruofei Zhong Cankun Yang Yunwei Tang Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures Remote Sensing satellite videos moving airplane detection moving airplane tracking template matching similarity measures |
| title | Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures |
| title_full | Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures |
| title_fullStr | Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures |
| title_full_unstemmed | Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures |
| title_short | Detecting and Tracking Moving Airplanes from Space Based on Normalized Frame Difference Labeling and Improved Similarity Measures |
| title_sort | detecting and tracking moving airplanes from space based on normalized frame difference labeling and improved similarity measures |
| topic | satellite videos moving airplane detection moving airplane tracking template matching similarity measures |
| url | https://www.mdpi.com/2072-4292/12/21/3589 |
| work_keys_str_mv | AT fanshi detectingandtrackingmovingairplanesfromspacebasedonnormalizedframedifferencelabelingandimprovedsimilaritymeasures AT fangqiu detectingandtrackingmovingairplanesfromspacebasedonnormalizedframedifferencelabelingandimprovedsimilaritymeasures AT xiaoli detectingandtrackingmovingairplanesfromspacebasedonnormalizedframedifferencelabelingandimprovedsimilaritymeasures AT ruofeizhong detectingandtrackingmovingairplanesfromspacebasedonnormalizedframedifferencelabelingandimprovedsimilaritymeasures AT cankunyang detectingandtrackingmovingairplanesfromspacebasedonnormalizedframedifferencelabelingandimprovedsimilaritymeasures AT yunweitang detectingandtrackingmovingairplanesfromspacebasedonnormalizedframedifferencelabelingandimprovedsimilaritymeasures |