Fig. 1. Location of research distribution (a)Independent sea ice distribution between narrow waterways bordering land,(b)independent sea ice distribution under thin ice disturbance,(c)normal independent sea ice distribution,(d)independent sea ice distribution under thin cloud disturbance
Fig. 2. Architecture of U-ASPP-Net
Fig. 3. Architecture of ASPP
Fig. 4. Convolution process of atrous depthwise separable convolution
Fig. 5. Schematic diagram of overlapping edge elimination strategy
Fig. 6. Schematic diagram of experimental data (a)schematic of data,(b)schematic of label
Fig. 7. Thumbnails of test set (a)Normal independent sea ice distribution, (b)independent sea ice distribution under thin ice disturbance, (c-d)independent sea ice distribution under thin cloud disturbance, (e)independent sea ice distribution between narrow waterways bordering land
Fig. 8. The changing trend of each indicator under different overlaps
Fig. 9. Independent sea ice segmentation results by different methods (a)Images to be segmented, (b)manually interpreted images, (c)U-ASPP-Net, (d)U-Net, (e)Deeplab v3+, (f)partition gradient difference and bimodal threshold segmentation algorithm
Fig. 10. Comparison of ability to resist interference of thin cloud and sea ice (a)Images to be segmented, (b)Manually interpreted images, (c)U-ASPP-Net, (d)U-Net, (e)Deeplab v3+, (f)partition gradient difference and bimodal threshold segmentation algorithm
性能参数 | 指标 |
---|
量化等级 | 12 bit | 扫描范围 | ±55.1°±0.1° | 每条扫描线采样点数 | 2 048(1 000 m),8 192(250 m) | 波段范围 | 0.4∼12.5 µm | 通道间像元配准 | <0.3个像元 | 定标精度 | 可见光和近红外通道:5%(反射率),星上定标器实现可见光星上定标(相对和绝对辐射)。 红外通道(星上黑体):0.5 K(270 K),分裂窗两个通道定标误差不一致性<小于0.5 K |
|
Table 1. Performance indicators of MERSI-Ⅱ
通道号 | 对应传感器通道号 | 中心波长 | 光谱带宽 | 空间分辨率 |
---|
1 | 1 | 0.470 | 0.05 | 250 | 2 | 2 | 0.550 | 0.05 | 250 | 3 | 3 | 0.650 | 0.05 | 250 | 4 | 4 | 0.865 | 0.05 | 250 | 5 | 6 | 1.640 | 0.02 | 250(重采样) |
|
Table 2. Channel characteristics of sea ice datasets
样本大小 | 192×192 |
---|
训练样本数量 | 1622幅 | 验证样本数量 | 302幅 | 独立海冰占比 | 0.19 |
|
Table 3. Distribution of independent sea ice datasets
混淆矩阵 | 真实值 |
---|
Positive | Negative |
---|
预测值 | Positive | TP | FP | Negative | FN | TN |
|
Table 4. Confusion matrix
损失函数 | OA | Kappa系数 | IOU | Dice系数 |
---|
FDWloss | 93.61% | 0.74 | 0.66 | 0.78 | Diceloss | 89.69% | 0.55 | 0.45 | 0.61 | Focalloss | 91.43% | 0.62 | 0.52 | 0.67 |
|
Table 5. Comparison of independent sea ice segmentation accuracies of U-ASPP-NET model based on different loss function
| 准确率 | Kappa | IOU | Dice系数 |
---|
U-ASPP-Net | 93.61% | 0.74 | 0.66 | 0.78 | U-Net | 91.20% | 0.68 | 0.58 | 0.73 | DeeplabV3+ | 90.65% | 0.59 | 0.48 | 0.65 | 分区梯度差分与双峰阈值分割法 | 86.07% | 0.40 | 0.31 | 0.47 |
|
Table 6. Comparison of accuracies for segmentation of independent sea ice with different methods