The study on multi sensors observing for ground truth using vehicle
Fig.5 shows averaged spectral reflectance of each observation site. Difference of vegetation density is appeared more distinctive reflectance in near infrared region than in red region.Fig. 6 shows 2D plot of convolved broad band reflectance in red and near infrared region. This figure is also indicated that difference of reflectance in near infrared is large. Moreover it is appeared that reflectance acquired in Site 2 and Site 3 has similar characteristics of soil reflectance, therefore it is necessary to investigate soil line of each site.
Fig.7 shows example of calculation result of VCR. Right image is original image Obtained by DSC, middle image is VC image derived from distance data (grass height data) and laser intensity data obtained by 3D scanner , left image is VC image calculated with RGB pattern method. RGB pattern method is able to calculate vegetation coverage using digital image lump together. But,in this study threshold is decided by above 3D scanner data. Comparing with original image and processed image, it is clear to acquire proper result. Fig. 8 shows calculated VCR at each measurement point. This figure indicates that more dense vegetated area has larger variable than sparse vegetated area and it could also acquire valid results. But above this method is still needed decision of threshold empirically. Therefore it is necessary to calculate VCR from so many digital images lump together complete automatically.
Fig.4 Ground temperature at each measurement points.
Fig.5 Averaged spectral reflectance in each observation site.
Fig. 6 2D plot of broad band reflectance in red ( 580nm-680nm) versus near infrared (725nm-620nm) region.
Fig. 7 Example of calculation result of VCR.
5.Discusssion
In this study, we developed new ground truth measurement system mounted multi sensor using vehicle. This system has advantage that most measurement operation can be conducted inside of vehicle. Therefore we could obtain various kinds of so many data in the short time with satellite scale. And we confirmed that data obtained by this system has potential for comparable with satellite data such as NOAA AVHRR.
Analysis after observation become more efficient and easier than previous measurement system because positioning data and the accuracy of GPS and the other information with spectral information is recorded in one ASCII file. We tested to calculate of VCR using 3D Scanner data and DSC image, and we confirmed it is very useful, but it is needed to examine for methodology of calculation of VCR.
Fig.8 Calculated VCR at each measurement points.
In future, it is thought that biomass data and so on obtained with in the same observation period is analyzed and it is estimated relationship between spectral reflectance and vegetation amount. And it should be also clarified representative value in each site. Moreover we plan to apply the ground-based data obtained by our new developed system to NOAA AVHRR LAC data.
6. Acknowledgements
This work has been supported by CREST (Core Research for Evolutional Science and Technology) of Japan Science and Technology Corporation (JST). And we are very grateful to many staffs of Dundgobi local government and The Information and Computer Center in Mongolia.
References
- Yamamoto, H., Kajiwara, K., Honda, Y.: Car Survey Measurement in Mongolia Grassland, Proceeding of the Symposium of the Japan Society of Photogrammetry and Remote Sensing, 241-244,1995.(in Japanese)
- Yamamoto,H., Kajiwara, K., Honda, Y.: The estimate of Vegetation Coverage Ratio by Car Survey ---- The Experience in Mongolia Grassland ---, Proceeding of the Symposium of the Japan Society of Photogrammetry and Remote Sensing, 241-244,1996.(in Japanese)
-
Yamamoto, H., Honda, Y., Kajiwara, K., Okano, C., Huang, S. and Matsuoka, .: The Analysis of Field Data Measured in Grasslandfor Biomass Monitoring Using Satellite Data, Journal of the Japan Society of Photogrammetry and Remote Sensing , Vol.36, No:4, 1997.(in Japanese)
