Evaluation of ground, proximal and aerial remote sensing technologies for crop stress monitoring. Issue 16 (2016)

• Record Type: Journal Article
• Authors: Zhou, Jianfeng; Khot, Lav R.; Bahlol, Haitham Y.; Boydston, Rick; Miklas, Phillip N.
• Is Part Of: IFAC-PapersOnLine. Volume 49:Issue 16(2016)
• Page Start: 22
• Page End: 26
• Publication Date: 2016
• Subjects: Remote sensing -- ground sensor -- sensing altitude -- vegetation index -- stress detection -- row crops; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2016.10.005
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: In this study, three different sensing technologies were evaluated for their performance in monitoring pinto beans crop stress at early stages. Treatments involved replicate pinto bean field plots with 50% and 100% irrigation throughout the season. Eight different pinto bean cultivars were seeded on the plots prepared with either strip or conventional tillage method. Evaluated technologies were a handheld linear ceptometer, and multi-spectral proximal and aerial remote sensing technologies. Spatial resolutions of the aerial remote sensing images acquired from 100 m above ground level (AGL) and the proximal sensing images acquired at 6.7 m AGL were 35.2 and 5.6 mm·pixel -1, respectively. Crop indictors of leaf area index (LAI), green normalized difference vegetation index (GNDVI) and canopy cover (CC) were extracted from the data of ceptometer and multispectral sensors collected at the early stages of pinto beans on July of 2015. Results show that spatial coverage of aerial remote sensing was thus 700 times larger than that of proximal remote sensing utilized in this study. GNDVI and CC data from both aerial and proximal remote sensing was able to discriminate crops with different irrigation and tillage treatment significantly at 5% level. Similarly, leaf area index (LAI) from ground sensor (ceptometer) was also able to distinguish effects of different irrigations, but could not differentiate tillage treatments. Correlation trends showed that the aerial remote sensing and ground sensing based indicators were strongly related with crop yield compared to proximal remote sensing based indicators. Although data were collected for natural light variations, possibly latter sensing module had more predominant light variation effect on image quality at different imaging times on given imaging day.