Fig. 1. Convolution kernel with different dilation rates. (a) r=1; (b) r=6; (c) r=12; (d) r=18

Fig. 2. Schematic of deep multiscale dilation fully convolutional neural network architecture

Fig. 3. Visual analysis on cloud area for different pooling layers

Fig. 4. Diagram of edge protection of training image

Fig. 5. Comparison of cloud detection results at different areas using different algorithms. (a) Area covered by thick cloud; (b) area covered by middle-thick cloud; (c) area covered by thick and thin clouds; (d) area covered by thick cloud and haze with the a complex scene; (e) area covered by a large range of thick cloud

Table 1. Multi-spectral image parameters of ZY-3 satellite

Table 2. Comparison of the accuracy of different network structures

Table 3. Quantity evaluation parameters of different algorithms