Fig. 1. Basic structure of FCN

Fig. 2. Diagram of network structure

Fig. 3. Diagram of regional consolidation

Fig. 4. Flow chart of classification method

Fig. 5. Original images and tag data examples. (a) Example 1; (b) example 2

Fig. 6. Classification results of different methods. (a) Original image; (b) segmentation result of mean-shift ①;(c) segmentation result of mean-shift ②; (d) segmentation result of mean-shift ③; (e) true classification image;(f) classification result of SVM ; (g) classification result of ANN; (h) classification result of FCN-16; (i) classification result of FCN-8; (j) classification result of proposed FCN; (k) classification result of proposed FCN adding segmentation result of mean-shift ①; (l) classifi

Fig. 7. Marked images of some details. (a) True classification image; (b) classification result of FCN-16; (c) classification result of proposed FCN adding segmentation result of mean-shift ②

Table 1. Confusion matrice and overall accuracy of SVM classification method%

Table 2. Confusion matrice and overall accuracy of ANN classification method%

Table 3. Confusion matrice and overall accuracy of FCN-16 classification method%

Table 4. Confusion matrice and overall accuracy of FCN-8 classification method%

Table 5. Confusion matrice and overall accuracy of proposed FCN classification method%

Table 6. Confusion matrice and overall accuracy of proposed FCN adding segmentation result of mean-shift ①%

Table 7. Confusion matrice and overall accuracy of proposed FCN adding segmentation result of mean-shift ②%

Table 8. Confusion matrice and overall accuracy of proposedFCN adding segmentation result of mean-shift ③%