GreenSeeker™ (Red NDVI) Performance Review Testing Summary |
|||||||||||||||||||||||||||||||||||||||||||||||||||||
GreenSeeker™ (Red NDVI) Performance Review, Testing Summary | |||||||||||||||||||||||||||||||||||||||||||||||||||||
GreenSeeker™ (Red NDVI) Performance Review, Testing Summary (Word FILE) | |||||||||||||||||||||||||||||||||||||||||||||||||||||
Objective: · To determine performance characteristics of GreenSeeker™ Rev. G (Red NDVI) sensors Testing Parameters: · All tests were conducted under steady state conditions unless otherwise noted. At least two minutes were allowed for the sensor to reach steady state before testing commenced. 1. Step Test - dynamic response · Target description: Teal and white strip target · Height held constant at 1 m from target to photodiode window · Target moved from teal strip to white strip to teal strip, velocity = 0.08m/sec · NDVI changes as sensor moves from teal area to white area, then to teal area again · Transition zone distance = 2.7 cm
Figure 1. Dynamic Response of GreenSeeker Sensor · Conclusion: Sensor responded appropriately with little ramping effect due to target changes. 2. Time Effect on Sensor Signal · Sensor was mounted stationary over a teal and white strip target · Height was held constant at 1 m from target to photodiode window · Test was conducted for two minutes
Figure 2. Two minute elapsed time test of GreenSeeker sensor· Average NDVI = 0.816, standard deviation in NDVI = 0.0056 · Conclusion: Sensor variation in NDVI throughout the two-minute test was small. 3. Temperature Test · Tested at 7°C, 21°C, and 42°C · At least two minutes allowed for sensor to reach steady state at each temperature · Three heights (0.94 m, 1 m, and 1.3 m) tested at each temperature · Teal strip target
Table 1. Variation due to temperature of GreenSeeker sensor
· Results: Temperature has little effect on NDVI within the specified temperature range. 4. Sensor-to-target Distance Test · Tested at 21°C · Teal strip target 0.91 m wide
Figure 3. NDVI change with change in GreenSeeker height
· Conclusion: NDVI is independent of height over the range tested (Slope = 0.0003, R2 = 0.0542) · Standard deviation at each height is shown in Figure 4
Figure 4. Standard deviation of NDVI with respect to change in sensor height· Conclusion: Variability of data increased slightly as height increased. However, the average NDVI was approximately the same. 5. Sensitivity across field of view · Tested 5 cm white plate on teal target and 5 cm teal plate on white target · 5 cm plates moved across target strip in 2.54 cm increments · Height held constant at 0.9 m from target to photodiode window · Tested at 21°C · Tested over the designed field of view of the WeedSeeker/GreenSeeker system
Figure 5. Sensitivity to white plate on teal strip background across GreenSeeker sensor field of view
Figure 6. Sensitivity to teal plate on white strip background across GreenSeeker sensor field of view· Conclusions: NDVI varied less than 8% across the field of view in the white plate on teal background test and 12% in the teal plate on white background test. 6. Sunlight effect · Height was held constant at 1.0 m from target to photodiode window · Various turf targets were used · Data was gathered in direct sunlight and with sunlight blocked with black 3 mil plastic cover
Table 2. Effect of sunlight on GreenSeeker sensor · Conclusion: Sunlight had little effect on NDVI (Mean difference = 0.00012 NDVI) 7. Plant-on-soil height test · Tested at 25°C, outdoors · Sensor positioned directly above a single plant (approx. 15 cm diameter) with a clear soil background. · Clear soil NDVI = 0.100
Figure 3. NDVI change with change in GreenSeeker height· Conclusion: Sensor can detect foliage with little variation in NDVI over a range of distances, 0.81 m to 1.22 m (32 in. to 48 in.) from canopy to photodiode window.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||