Y distinguish and recognize vegetation and bare land. Figure 4 four shows the
Y distinguish and identify vegetation and bare land. Figure 4 4 shows the difference amongst vegetation and Figure shows the difference involving vegetation and determine vegetation and bare land. bare land inin the NDVIfragment segment. the NDVI fragment segment. bare landFigure 4. Trajectory datadata and corresponding images false colorcomposite image (SWIR2/NIR/Green). The cross cross shape is Figure four. Trajectory and corresponding pictures false color composite image (SWIR2/NIR/Green). The shape would be the place in the sample point; black point may be the original NDVI value; the green line is fitted vegetation track; the the place in the sample point; thethe black point is theoriginal NDVI worth; the green line is the the fitted vegetation track; the blue line may be the fitted bare ground track; the red point is definitely the breakpoint of vegetation damaged; and the orange point is the breakpoint of vegetation reclamation.two.5. Identification of Damage and Reclamation Spatio-Temporal Processing In the method of mining, the surface is stripped or covered by slag, resulting inside a sharp decline in vegetation coverage. Firstly, we fit and segment the NDVI trajectory to have the NDVI segmentation by the CCDC algorithm. Then the disturbance pixels of vegetation are extracted by the transform amplitude of NDVI fragments. Reference He et al. (2020) [29] the process for determining the threshold, we select 100 harm (60 reclamation) sample points plus the parameter in [0.2, 0.6] ([-0.2, -0.6]) by the interval of 0.05 to calculate the Elsulfavirine Inhibitor accuracy of detecting damage (reclamation). Ultimately, we choose the reduce (enhance) of NDVI by 0.3 (0.25) as the optimal threshold to identify damage (reclamation). For multiple-segmented pixels, the minimum of a number of harm time is set because the final harm time, although the maximum of numerous reclamation time is set as the final reclamation time. Importantly, the end from the trajectory must be an ascending segment. Consequently, the broken and reclaimed time mapping with the region is completed. To cut down the noise in the patches of damaged time, the broken time in the adjacent pixels is mainly continuous. Hence, we smooth the damaged-time patches by mode algorithm. It really is worth noting that, for the pixels which have been broken before 1986 and have not been reclaimed for the duration of the study period, we set the damaged time of these to 1 January 1986. two.6. Validation Tachysterol 3 Purity & Documentation Considering the difficulty to get public remote-sensing data having a high time-andspatial resolution, we verify the accuracy of abrupt transform time per year. Fifty points per year are randomly selected within the broken location, though twenty points per year are selected inside the reclamation area. The detection time of damage year is from 1986 to 2020, and that of reclamation year is from 1988 to 2020. 1750 harm samples and 660 reclamation samples were detected. Then, the high-resolution image data on Google Earth are applied for interactive visual calibration to figure out the harm year and recovery year of each sampling point. By comparing the sample label with all the recognition outcomes of the algorithm,Remote Sens. 2021, 13,8 ofthe user accuracy, producer accuracy, all round accuracy, and kappa coefficient of mining damage and reclamation detection are calculated, and also the alter detection and accuracy are verified. 3. Benefits three.1. Accuracy The all round accuracy of harm and recovery is 92 and 88 respectively, while the kappa coefficients are 85 and 84 respectively (Figure 5).
