Scholarlink

Purpose <p>This study aimed at comparing and correlating six vegetation indices (VI) derived from an Unmanned Aerial Vehicle (UAV) and two satellite platforms, PlanetScope (PS) and Sentinel-2 (S2), in a cherry tomato (Solanum lycopersicum) crop.</p> Methods <p>Multispectral images were acquired on seven dates during the crop cycle, extracting the mean VI values for two treatment plots. An ANOVA was conducted, and Pearson’s correlation coefficient (r) was computed.</p> Results <p>The ANOVA confirmed that VI values from the UAV were statistically distinct from satellite platforms (<i>p</i> &lt; 0.05). Despite this difference, very strong linear correlations were found, particularly between the UAV and Sentinel-2, with Pearson coefficients (r) of 0.957 (NDVI) and 0.958 (OSAVI). This strong relationship enabled the generation of robust linear regression models to serve as cross-calibration algorithms. These models demonstrated high predictive power, with coefficients of determination (R²) reaching 0.916 for the UAV/S2 comparison of VIs (NDVI, IPVI, and OSAVI).</p> Conclusion <p>This approach allows to merge information from UAVs and satellites, yielding consistent datasets that improve the accuracy of crop monitoring. These results show that while UAV-based measurements are different, satellite platforms reliably capture the spatial variation in crop health, thereby confirming their utility as scalable instruments for agricultural monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Crop Monitoring with Multiple Sensors: A Comparative Analysis and Validation of UAV, PlanetScope, and Sentinel-2 in Cherry Tomato

  • 期刊论文

  • 发表日期：

    2026年3月4日

查看全文
  • Osiris Chávez-Martínez,
  • Sergio Alberto Monjardin-Armenta,
  • Jesús Gabriel Rangel-Peraza,
  • Zuriel Dathan Mora-Félix,
  • Antonio Jesús Sanhouse-García

摘要

Purpose

This study aimed at comparing and correlating six vegetation indices (VI) derived from an Unmanned Aerial Vehicle (UAV) and two satellite platforms, PlanetScope (PS) and Sentinel-2 (S2), in a cherry tomato (Solanum lycopersicum) crop.

Methods

Multispectral images were acquired on seven dates during the crop cycle, extracting the mean VI values for two treatment plots. An ANOVA was conducted, and Pearson’s correlation coefficient (r) was computed.

Results

The ANOVA confirmed that VI values from the UAV were statistically distinct from satellite platforms (p < 0.05). Despite this difference, very strong linear correlations were found, particularly between the UAV and Sentinel-2, with Pearson coefficients (r) of 0.957 (NDVI) and 0.958 (OSAVI). This strong relationship enabled the generation of robust linear regression models to serve as cross-calibration algorithms. These models demonstrated high predictive power, with coefficients of determination (R²) reaching 0.916 for the UAV/S2 comparison of VIs (NDVI, IPVI, and OSAVI).

Conclusion

This approach allows to merge information from UAVs and satellites, yielding consistent datasets that improve the accuracy of crop monitoring. These results show that while UAV-based measurements are different, satellite platforms reliably capture the spatial variation in crop health, thereby confirming their utility as scalable instruments for agricultural monitoring.