Guang Chen
ABSTRACT
We propose computer vision based approach for effectively computing the vegetation coverage of the image to determine the structure of the vegetation and to understand wildlife habitat. To deal with the variation of lighting condition, two-stage segmentation strategy is applied. Firstly, texture information is used to roughly classify the vegetation and the reference blackboard at each position using Support Vector Machine. And then a K-means based adaptive color model is used to refine the segmentation result in pixel level. We evaluate our approach on our dataset, and the results demonstrate that the proposed method is robust to environment changing, and color instability. For blackboard localization, we tested 200 images and the accuracy is approximately 93%. For grass detection and coverage computation, the error rate is approximately 3%.
- Jorgensen, C.F., Stutzman, R.J., Anderson, L.C., E Decker, S., Powell, L.A., Schacht, W.H. and Fontaine, J.J. 2013. Choosing a DIVA: a comparison of emerging digital imagery vegetation analysis techniques. Applied Vegetation Science, 16(4), pp. 552--560.Google Scholar
Cross Ref
- Limb, R.F., Hickman, K.R., Engle, D.M., Norland, J.E. and Fuhlendorf, S.D. 2007. Digital photography: reduced investigator variation in visual obstruction measurements for southern tallgrass prairie. Rangeland Ecology & Management, 60(5), pp. 548--552.Google ScholarCross Ref
- Leis, S.A. and Morrison, L.W. 2011. Field test of digital photography biomass estimation technique in tallgrass prairie. Rangeland ecology & management, 64(1), pp. 99--103.Google Scholar
- Morrison, L.W. 2015. Observer error in vegetation surveys: a review. Journal of Plant Ecology, 9(4), pp. 367--379.Google ScholarCross Ref
- Robel, R.J., Briggs, J.N., Dayton, A.D. and Hulbert, L.C. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management, pp. 295--297.Google ScholarCross Ref
- Nudds, T.D. 1977. Quantifying the vegetative structure of wildlife cover. Wildlife Society Bulletin, pp. 113--117.Google Scholar
- Cagney, J., Cox, S.E. and Booth, D.T. 2011. Comparison of point intercept and image analysis for monitoring rangeland transects. Rangeland Ecology & Management, 64(3), pp. 309--315.Google ScholarCross Ref
- Booth, D.T., Cox, S.E. and Johnson, D.E. 2005. Detection-threshold calibration and other factors influencing digital measurements of ground cover. Rangeland Ecology & Management, 58(6), pp. 598--604.Google ScholarCross Ref
- Dalal, N. and Triggs, B. 2005, June. Histograms of oriented gradients for human detection. In Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on (Vol. 1, pp. 886--893). IEEE. Google ScholarDigital Library
- He, D.C. and Wang, L. 1990. Texture unit, texture spectrum, and texture analysis. IEEE transactions on Geoscience and Remote Sensing, 28(4), pp. 509--512.Google Scholar
Index Terms
- Digital Image Vegetation Analysis with Machine Learning
Recommendations
Variation Characteristics Analysis of the Vegetation Coverage in Midu County Based on Landsat 8 Remote Sensing Image
ICIIP '17: Proceedings of the 2nd International Conference on Intelligent Information ProcessingThe purpose of this study is to analyze the general characteristics of the spatial and temporal changes of vegetation coverage of Midu County in 2013, 2014 and 2015. Based on the Landsat 8 satellite remote sensing image data of Midu County in the three ...
Temporal and Spatial Change Monitoring of Vegetation Coverage in Haiyan County Based on GF-1
AbstractBased on the GF-1 satellite remote sensing data of the forage growing season in 2017 and 2018, the vegetation coverage was obtained by using dimidiate pixel model in Haiyan County demonstration area. The study area was divided into three areas ...
Temporal and Spatial Changes Monitoring of Vegetation Coverage in Qilian County Based on GF-1 Image
AbstractQilian County demonstration area is an important ecological security barrier in the country. Based on the GF-1 image, the dimidiate pixel model was used to invert vegetation coverage in Qilian County demonstration area. The temporal and spatial ...
Comments