Sugarcane canopy detection using high spatial resolution UAS images and digital surface model
The use of an unmanned aerial system (UAS) equipped with multispectral cameras is a potential approach to acquire canopy reflectance to make various correlations with desired crop parameters. However, the acquired reflectance data are mixed with unwanted data, such as reflectance from soil, which si...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Khon Kaen University
2019-12-01
|
| Series: | Engineering and Applied Science Research |
| Subjects: | |
| Online Access: | https://www.tci-thaijo.org/index.php/easr/article/download/192497/155870/ |
| _version_ | 1811264585720659968 |
|---|---|
| author | Chanreaksa Chea Khwantri Saengprachatanarug Jetsada Posom Mahisorn Wongphati Eizo Taira |
| author_facet | Chanreaksa Chea Khwantri Saengprachatanarug Jetsada Posom Mahisorn Wongphati Eizo Taira |
| author_sort | Chanreaksa Chea |
| collection | DOAJ |
| description | The use of an unmanned aerial system (UAS) equipped with multispectral cameras is a potential approach to acquire canopy reflectance to make various correlations with desired crop parameters. However, the acquired reflectance data are mixed with unwanted data, such as reflectance from soil, which significantly affects some commonly used vegetation indices, such as the NDVI. This study compares the performance of three methods for detecting the canopy area of 3-month-old sugarcane crops. These methods extract the canopy areas using 5 NDVI thresholds (0.2, 0.3, 0.4, 0.5, and 0.6), a principal component analysis (PCA) threshold, and a digital surface model (DSM) threshold. The performance assessment will deliberately consider the quality percentage (QP) of each method to correctly detect the canopy area of short sugarcane crops in 10 selected images. The results show that filtration by the PCA threshold method provides the best result with a QP of 65.89-78.72%. The NDVI threshold method at levels of 0.3 and 0.4 follow with QPs of 58.42-68.81% and 40.80-70.81%, respectively, and the lowest accuracy is obtained by the DSM threshold method, which has QPs of 14.80-30.78%. |
| first_indexed | 2024-04-12T20:07:00Z |
| format | Article |
| id | doaj.art-74c387b73fa9451baf221992beb681c1 |
| institution | Directory Open Access Journal |
| issn | 2539-6161 2539-6218 |
| language | English |
| last_indexed | 2024-04-12T20:07:00Z |
| publishDate | 2019-12-01 |
| publisher | Khon Kaen University |
| record_format | Article |
| series | Engineering and Applied Science Research |
| spelling | doaj.art-74c387b73fa9451baf221992beb681c12022-12-22T03:18:22ZengKhon Kaen UniversityEngineering and Applied Science Research2539-61612539-62182019-12-0146431231710.14456/easr.2019.35Sugarcane canopy detection using high spatial resolution UAS images and digital surface modelChanreaksa CheaKhwantri SaengprachatanarugJetsada PosomMahisorn WongphatiEizo TairaThe use of an unmanned aerial system (UAS) equipped with multispectral cameras is a potential approach to acquire canopy reflectance to make various correlations with desired crop parameters. However, the acquired reflectance data are mixed with unwanted data, such as reflectance from soil, which significantly affects some commonly used vegetation indices, such as the NDVI. This study compares the performance of three methods for detecting the canopy area of 3-month-old sugarcane crops. These methods extract the canopy areas using 5 NDVI thresholds (0.2, 0.3, 0.4, 0.5, and 0.6), a principal component analysis (PCA) threshold, and a digital surface model (DSM) threshold. The performance assessment will deliberately consider the quality percentage (QP) of each method to correctly detect the canopy area of short sugarcane crops in 10 selected images. The results show that filtration by the PCA threshold method provides the best result with a QP of 65.89-78.72%. The NDVI threshold method at levels of 0.3 and 0.4 follow with QPs of 58.42-68.81% and 40.80-70.81%, respectively, and the lowest accuracy is obtained by the DSM threshold method, which has QPs of 14.80-30.78%.https://www.tci-thaijo.org/index.php/easr/article/download/192497/155870/canopy detectionunmanned aerial system (uas)digital surface modelprincipal component analysisnormalized vegetation indexmultispectral image |
| spellingShingle | Chanreaksa Chea Khwantri Saengprachatanarug Jetsada Posom Mahisorn Wongphati Eizo Taira Sugarcane canopy detection using high spatial resolution UAS images and digital surface model Engineering and Applied Science Research canopy detection unmanned aerial system (uas) digital surface model principal component analysis normalized vegetation index multispectral image |
| title | Sugarcane canopy detection using high spatial resolution UAS images and digital surface model |
| title_full | Sugarcane canopy detection using high spatial resolution UAS images and digital surface model |
| title_fullStr | Sugarcane canopy detection using high spatial resolution UAS images and digital surface model |
| title_full_unstemmed | Sugarcane canopy detection using high spatial resolution UAS images and digital surface model |
| title_short | Sugarcane canopy detection using high spatial resolution UAS images and digital surface model |
| title_sort | sugarcane canopy detection using high spatial resolution uas images and digital surface model |
| topic | canopy detection unmanned aerial system (uas) digital surface model principal component analysis normalized vegetation index multispectral image |
| url | https://www.tci-thaijo.org/index.php/easr/article/download/192497/155870/ |
| work_keys_str_mv | AT chanreaksachea sugarcanecanopydetectionusinghighspatialresolutionuasimagesanddigitalsurfacemodel AT khwantrisaengprachatanarug sugarcanecanopydetectionusinghighspatialresolutionuasimagesanddigitalsurfacemodel AT jetsadaposom sugarcanecanopydetectionusinghighspatialresolutionuasimagesanddigitalsurfacemodel AT mahisornwongphati sugarcanecanopydetectionusinghighspatialresolutionuasimagesanddigitalsurfacemodel AT eizotaira sugarcanecanopydetectionusinghighspatialresolutionuasimagesanddigitalsurfacemodel |