A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121
<i>Papaver somniferum</i> (opium poppy) is not only a source of raw material for the production of medical narcotic analgesics but also the major raw material for certain psychotropic drugs. Therefore, it is stipulated by law that the cultivation of <i>Papaver somniferum</i>...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-04-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/8/1834 |
| _version_ | 1797443908874010624 |
|---|---|
| author | Qian Wang Chunshan Wang Huarui Wu Chunjiang Zhao Guifa Teng Yajie Yu Huaji Zhu |
| author_facet | Qian Wang Chunshan Wang Huarui Wu Chunjiang Zhao Guifa Teng Yajie Yu Huaji Zhu |
| author_sort | Qian Wang |
| collection | DOAJ |
| description | <i>Papaver somniferum</i> (opium poppy) is not only a source of raw material for the production of medical narcotic analgesics but also the major raw material for certain psychotropic drugs. Therefore, it is stipulated by law that the cultivation of <i>Papaver somniferum</i> must be authorized by the government under stringent supervision. In certain areas, unauthorized and illicit <i>Papaver somniferum</i> cultivation on private-owned lands occurs from time to time. These illegal <i>Papaver somniferum</i> cultivation sites are dispersedly-distributed and highly-concealed, therefore becoming a tough problem for government supervision. The low-altitude inspection of <i>Papaver somniferum</i> cultivation by unmanned aerial vehicles has the advantages of high efficiency and time saving, but the large amount of image data collected needs to be manually screened, which not only consumes a lot of manpower and material resources but also easily causes omissions. In response to the above problems, this paper proposed a two-stage (target detection and image classification) method for the detection of <i>Papaver somniferum</i> cultivation sites. In the first stage, the YOLOv5s algorithm was used to detect <i>Papaver somniferum</i> images for the purpose of identifying all the suspicious <i>Papaver somniferum</i> images from the original data. In the second stage, the DenseNet121 network was used to classify the detection results from the first stage, so as to exclude the targets other than <i>Papaver somniferum</i> and retain the images containing <i>Papaver somniferum</i> only. For the first stage, YOLOv5s achieved the best overall performance among mainstream target detection models, with a Precision of 97.7%, Recall of 94.9%, and mAP of 97.4%. For the second stage, DenseNet121 with pre-training achieved the best overall performance, with a classification accuracy of 97.33% and a Precision of 95.81%. The experimental comparison results between the one-stage method and the two-stage method suggest that the Recall of the two methods remained the same, but the two-stage method reduced the number of falsely detected images by 73.88%, which greatly reduces the workload for subsequent manual screening of remote sensing <i>Papaver somniferum</i> images. The achievement of this paper provides an effective technical means to solve the problem in the supervision of illicit <i>Papaver somniferum</i> cultivation. |
| first_indexed | 2024-03-09T13:03:54Z |
| format | Article |
| id | doaj.art-853f8ba081ab4e70b417d8781e445079 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T13:03:54Z |
| publishDate | 2022-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-853f8ba081ab4e70b417d8781e4450792023-11-30T21:50:37ZengMDPI AGRemote Sensing2072-42922022-04-01148183410.3390/rs14081834A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121Qian Wang0Chunshan Wang1Huarui Wu2Chunjiang Zhao3Guifa Teng4Yajie Yu5Huaji Zhu6School of Information Science and Technology, Hebei Agricultural University, Baoding 071001, ChinaSchool of Information Science and Technology, Hebei Agricultural University, Baoding 071001, ChinaNational Engineering Research Center for Information Technology in Agriculture, Beijing 100097, ChinaNational Engineering Research Center for Information Technology in Agriculture, Beijing 100097, ChinaSchool of Information Science and Technology, Hebei Agricultural University, Baoding 071001, ChinaSchool of Software, Tsinghua University, Beijing 100084, ChinaNational Engineering Research Center for Information Technology in Agriculture, Beijing 100097, China<i>Papaver somniferum</i> (opium poppy) is not only a source of raw material for the production of medical narcotic analgesics but also the major raw material for certain psychotropic drugs. Therefore, it is stipulated by law that the cultivation of <i>Papaver somniferum</i> must be authorized by the government under stringent supervision. In certain areas, unauthorized and illicit <i>Papaver somniferum</i> cultivation on private-owned lands occurs from time to time. These illegal <i>Papaver somniferum</i> cultivation sites are dispersedly-distributed and highly-concealed, therefore becoming a tough problem for government supervision. The low-altitude inspection of <i>Papaver somniferum</i> cultivation by unmanned aerial vehicles has the advantages of high efficiency and time saving, but the large amount of image data collected needs to be manually screened, which not only consumes a lot of manpower and material resources but also easily causes omissions. In response to the above problems, this paper proposed a two-stage (target detection and image classification) method for the detection of <i>Papaver somniferum</i> cultivation sites. In the first stage, the YOLOv5s algorithm was used to detect <i>Papaver somniferum</i> images for the purpose of identifying all the suspicious <i>Papaver somniferum</i> images from the original data. In the second stage, the DenseNet121 network was used to classify the detection results from the first stage, so as to exclude the targets other than <i>Papaver somniferum</i> and retain the images containing <i>Papaver somniferum</i> only. For the first stage, YOLOv5s achieved the best overall performance among mainstream target detection models, with a Precision of 97.7%, Recall of 94.9%, and mAP of 97.4%. For the second stage, DenseNet121 with pre-training achieved the best overall performance, with a classification accuracy of 97.33% and a Precision of 95.81%. The experimental comparison results between the one-stage method and the two-stage method suggest that the Recall of the two methods remained the same, but the two-stage method reduced the number of falsely detected images by 73.88%, which greatly reduces the workload for subsequent manual screening of remote sensing <i>Papaver somniferum</i> images. The achievement of this paper provides an effective technical means to solve the problem in the supervision of illicit <i>Papaver somniferum</i> cultivation.https://www.mdpi.com/2072-4292/14/8/1834<i>Papaver somniferum</i> inspectionunmanned aerial vehicle (UAV)small target detectionYOLOv5stwo-stage detection and classification |
| spellingShingle | Qian Wang Chunshan Wang Huarui Wu Chunjiang Zhao Guifa Teng Yajie Yu Huaji Zhu A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121 Remote Sensing <i>Papaver somniferum</i> inspection unmanned aerial vehicle (UAV) small target detection YOLOv5s two-stage detection and classification |
| title | A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121 |
| title_full | A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121 |
| title_fullStr | A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121 |
| title_full_unstemmed | A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121 |
| title_short | A Two-Stage Low-Altitude Remote Sensing Papaver Somniferum Image Detection System Based on YOLOv5s+DenseNet121 |
| title_sort | two stage low altitude remote sensing papaver somniferum image detection system based on yolov5s densenet121 |
| topic | <i>Papaver somniferum</i> inspection unmanned aerial vehicle (UAV) small target detection YOLOv5s two-stage detection and classification |
| url | https://www.mdpi.com/2072-4292/14/8/1834 |
| work_keys_str_mv | AT qianwang atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT chunshanwang atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT huaruiwu atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT chunjiangzhao atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT guifateng atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT yajieyu atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT huajizhu atwostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT qianwang twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT chunshanwang twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT huaruiwu twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT chunjiangzhao twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT guifateng twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT yajieyu twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 AT huajizhu twostagelowaltituderemotesensingpapaversomniferumimagedetectionsystembasedonyolov5sdensenet121 |