A novel method for extracting green fractional vegetation cover from digital images

Abstract

Question Although digital photography is an efficient and objective means of extracting green fractional vegetation cover (FVC), it lacks automation and classification accuracy. How can green FVC be extracted from digital images in an accurate and automated method? Methods Several colour spaces were compared on the basis of a separability index, and CIE L*a*b* was shown to be optimal for the tested colour spaces. Thus, all image processing was performed in CIE L*a*b* colour space. Gaussian models were used to fit the green vegetation and background distributions of the a* component. Three strategies (T0, T1 and T2 thresholding method) were tested to select the optimal thresholds for segmenting the image into green vegetation and non‐green vegetation. The a* components of the images were then segmented and the green FVC extracted. Results The FVC extracted using T0, T1, and T2 thresholding methods were evaluated with simulated images, and cross‐validated with FVC extracted with supervised classification methods. The results show that FVC extracted with T0, T1 and T2 thresholding methods are similar to those estimated with supervised classification methods. The mean errors associated with the FVC values provided in our approach and supervised classification are less than 0.035. In a test with simulated data, our method performed better than the supervised classification method. Conclusions Methods presented in this paper were demonstrated to be feasible and applicable for automatically and accurately extracting FVC of several green vegetation types with varying background and shadow conditions. However, our algorithm design assumes a Gaussian distribution for both vegetated and non‐vegetated portions of a digital image; moreover, the impact of view angle on the FVC extraction from digital images must also be considered.

Evaluation of Sampling Methods for Validation of Remotely Sensed Fractional Vegetation Cover

“…Details of the digital photography measurements can be found in reference [14,36]. The FVC of each photo was extracted using a published algorithm [37] that supposed that the green vegetation and background distributions of the greenness component in the color space were Gaussian; then, image The ground measurements were taken in 23 plots ( Figure 3). Each plot covers an area of 10 mˆ10 m in the cropland and 30 mˆ30 m in the fruit orchard and woodland.…”

“…Details of the digital photography measurements can be found in reference [14,36]. The FVC of each photo was extracted using a published algorithm [37] that supposed that the green vegetation and background distributions of the greenness component in the color space were Gaussian; then, image segmentation was performed based on this assumption. A previous study suggested that this method obtains a stable absolute error of less than 0.05 (with an FVC range of 0 to 1) [38].…”

Extracting Leaf Area Index by Sunlit Foliage Component from Downward-Looking Digital Photography under Clear-Sky Conditions

“…The CIE L*a*b* color space separates the luminance channel L* from two chromaticity channels a* and b*, which makes the correlations between channels minimal, and reduces the potential impact of illumination changes on image processing, compared with the RGB space. Negative a* indicates green, while positive a* indicates red, which has been implemented in previous studies to quantify the greenness of foliage [38,40].…”

A Robust Algorithm for Estimating Surface Fractional Vegetation Cover from Landsat Data

“…To remove distortion from center projection of the digital camera, the edges, about 40% of the length and width of the image, were cut off. Then, FVC values of the cropped images were extracted using the improved Gaussian simulation and segmentation method in CIE L*a*b* color space [60]. Finally, the FVC calculated from digital image were used to directly evaluate the performance of the proposed algorithm.…”

“…First, the image edges were cut off before FVC extraction [59] to remove distortion and perspective effects. Then, the thresholding method was used to extract FVC from the digital images [60]. This automatic classification method transfers images from the RGB color space to the Commission Internationale d'Eclairage LAB color space, which can easily distinguish green vegetation from the background.…”