Figures & data
Figure 3. Visualization of the extracted features. The first row illustrates four object categories in remote sensing images: farmland, town, river and road. The second row shows the features of the four objects learned by our method from the corresponding images in the first row
![Figure 3. Visualization of the extracted features. The first row illustrates four object categories in remote sensing images: farmland, town, river and road. The second row shows the features of the four objects learned by our method from the corresponding images in the first row](/cms/asset/a2c5f696-c40b-44fa-924a-5b328cbaa75d/tbed_a_1967259_f0003_c.jpg)
Figure 4. Average AGR-pcGDP of the NPCs in each quantile range and their respective non-NPCs. Error bars indicate 1 standard deviation across the counties
![Figure 4. Average AGR-pcGDP of the NPCs in each quantile range and their respective non-NPCs. Error bars indicate 1 standard deviation across the counties](/cms/asset/a179e216-db52-43b4-807a-ec1d0ba9c7ed/tbed_a_1967259_f0004_c.jpg)
Figure 5. Average AGR-pcGDPs for the NPCs and non-NPCs in the four provinces. Error bars indicate 1 standard deviation across the counties. The provincial growth rates data are obtained directly from the China Statistical Yearbook for comparison
![Figure 5. Average AGR-pcGDPs for the NPCs and non-NPCs in the four provinces. Error bars indicate 1 standard deviation across the counties. The provincial growth rates data are obtained directly from the China Statistical Yearbook for comparison](/cms/asset/45939553-22a9-4393-9075-7fcce95a44a1/tbed_a_1967259_f0005_c.jpg)
Supplemental Material
Download PDF (530.2 KB)Data availability statement
The Per capita GDP results and growth rates at different levels are available at https: 10.6084/m9.figshare.15052545 and https://www.doi.org/10.11922/sciencedb.j00076.00089.