Spectral information mining plays a significant role in tree species classification using remote sensing images but a measure of the importance of the mined features is also necessary. We created five new spectral indices and used MLC-RFI to determine their importance. The results showed that these spectral indices had a positive effect on tree species classification, and their importance was ranked as follows: PAI₅₇₈ > PAI₆₇₈ > TAI₅₇₈ > TAI₆₇₈ > SI₅₇. The experimental results showed that important spectral indices for tree species classification can be created and that it is beneficial to mine spectral information from remote sensing images.

The size of the texture extraction window impacts image tree species classification, and the determination of the optimal texture extraction window requires the supervision of a specific classifier for accuracy. Therefore, it is necessary to analyse which kind of classifier is more suitable and should be to choose. In this study, we extracted eight types of textures, namely mean, variance, homogeneity, contrast, dissimilarity, entropy, second moment and correlation, changed the window size by gradient increase and used maximum likelihood classification (MLC) and random forest (RF) to supervise and determine their optimal extraction windows, respectively. Finally, the optimised time consumption and classification accuracy for tree species classification was identified. The time consumption of MLC was significantly less than that of RF; however, neither was very long; for most textures, the optimal texture extraction window determined by MLC supervision was larger than that determined by RF supervision; in the classification of most feature sets, the overall accuracy obtained by MLC was less than that of RF. Because the time consumption of the texture extraction was much greater than that of the image classification, the comprehensive trade-off indicates that using RF supervision to determine the optimal window for texture extraction was more conducive to tree species recognition.

Vegetation coverage is an important indicator in regional ecological environment monitoring and plays a key role in its quality assessment. We consider Landsat7 ETM+ in 2000 and Landsat8 in 2016 as data sources using a different time phase partial image substitution method to eliminate cloud effects and an NDVI dimidiate pixel model to invert the vegetation coverage of the two time phases. We further classify them into five grades, provide statistics and analyse the areas of different grades at different time periods, while monitoring the spatial evolution of vegetation coverage over the past 16 years in Tongren. Experimental results showed that: (1) the different time phase partial image substitution method could reduce the influence of clouds on vegetation extraction; (2) in Tongren, the vegetation coverage area was decreased from 17,300.1 km² to 17,224.8 km² (i.e. decreased by 75.3 km²); (3) the areas of grade I and V increased by 0.42% and 15.08%, respectively, whereas the areas of grade II, III and IV decreased by 3.15%, 6.98% and 5.37%, respectively, which indicates that most of the area containing lower amount of vegetation gradually altered into an area containing a higher coverage of vegetation, whereas a few areas became bare land; and (4) the vegetation areas decreased due to expansion of cities and construction of dams, while vegetation increased due to the cultivation of crops and trees. Research shows that the overall evolution of vegetation coverage in Tongren is considerably good. However, while undertaking future development in the mountainous Karst region, one should be aware of the land’s intensive use and environmental protection.