Journal of Geo-information Science >
Scale Transformation Algorithm for Remote Sensing Imagery Based on Greatest Common Divisor
Received date: 2015-05-11
Request revised date: 2015-08-03
Online published: 2015-12-20
Copyright
The abundant remote sensing data with various spatial, radiational and spectral resolutions from multi-platforms provide rich information sources for the study of land surface information changes at different scales. Scale variation and sensitivity have a great impact on the application of remote sensing imagery in different scientific fields. We proposed a transformation algorithm to unify the scales for comparing data at different scales. The method is a scale transformation algorithm based on the greatest common divisor (STAGCD). Firstly, the greatest common divisor (GCD) between two different spatial scales is calculated. Secondly, according to the GCD, a GCD image will be produced by resampling the original remote sensing image. Finally, the new scale image will be obtained according to certain intervals for row and column to choose data from the GCD image. Several scale transformation algorithms have been employed in the test of the scale unification for an IKONOS image, including STAGCD and some other algorithms from professional software packages, such as ER Mapper, ERDAS, Matlab and so on. The effectiveness of these algorithms has been evaluated based on the information keeping degree compared with the original remote sensing image. A total of six indicators have been used for quantitative evaluation of the scale transformed images. The histogram and probability density function of Gauss based on kernel bandwidth optimization have been used for visual interpretation of the scale transformed images. The results show that the STAGCD image has adequate ability for keeping the information of original image. When scaling-down, STAGCD only increases the image size, but cannot improve the image’s spatial resolution. When scaling-up, STAGCD not only reduces the image size, but also decreases the image resolution. The STAGCD method is simple and can transform remote sensing imagery at different scales. The method provides an effective solution for the scale transformation between images without an integer multiple relationship.
GAO Yonggang* , XU Hanqiu . Scale Transformation Algorithm for Remote Sensing Imagery Based on Greatest Common Divisor[J]. Journal of Geo-information Science, 2015 , 17(12) : 1520 -1528 . DOI: 10.3724/SP.J.1047.2015.01520
Fig. 1 Flowchart of indirect comparison method图1 间接比较法流程图 |
Fig. 2 Flowchart of scale transformation experiment图2 尺度转换实验流程图 |
Fig. 3 Original and scale transformation imageries of IKONOS (Scale: 2.1 m)图3 IKONOS原始影像与尺度转换后影像(尺度:2.1 m) |
Fig. 4 Comparison diagram of histogram of original and scale transformation imageries (Scale: 2.1 m)图4 IKONOS原始影像与各尺度转换后影像直方图对比图(尺度:2.1 m) |
Tab. 1 Quantitative evaluation of the scale transformation algorithms (Scale: 2.1 m)表1 IKONOS原始影像与尺度转换后影像定量评价表(尺度: 2.1 m) |
方法 | 波段 | 评价指标 | |||||
---|---|---|---|---|---|---|---|
Mean | STD | SD | RMSE | ERGAS | CC | ||
原始影像 | 1 | 457.501 | 88.939 | - | - | - | - |
2 | 499.650 | 134.548 | - | - | - | - | |
3 | 378.310 | 153.395 | - | - | - | - | |
4 | 504.025 | 174.153 | - | - | - | - | |
商业(软件)算法 | 1 | 457.658 | 89.127 | 25.827 | 52.062 | 11.376 | 0.859 |
2 | 499.956 | 134.772 | 40.693 | 77.405 | 15.482 | 0.835 | |
3 | 378.582 | 153.520 | 46.246 | 85.474 | 22.577 | 0.845 | |
4 | 505.493 | 174.515 | 68.193 | 116.083 | 22.964 | 0.779 | |
本文算法 | 1 | 457.501 | 88.939 | 0.000 | 0.000 | 0.000 | 1.000 |
2 | 499.650 | 134.548 | 0.000 | 0.000 | 0.000 | 1.000 | |
3 | 378.310 | 153.395 | 0.000 | 0.000 | 0.000 | 1.000 | |
4 | 504.025 | 174.153 | 0.000 | 0.000 | 0.000 | 1.000 | |
商业(软件)算法→本文算法 | 1 | 457.324 | 88.954 | 4.622 | 19.211 | 4.201 | 0.977 |
2 | 499.254 | 134.478 | 7.528 | 30.255 | 6.060 | 0.974 | |
3 | 377.748 | 153.117 | 8.740 | 35.321 | 9.350 | 0.973 | |
4 | 503.760 | 174.071 | 12.328 | 49.902 | 9.906 | 0.959 | |
本文算法→商业(软件)算法 | 1 | 457.532 | 89.015 | 17.269 | 40.397 | 8.829 | 0.896 |
2 | 499.827 | 134.672 | 27.521 | 61.326 | 12.272 | 0.895 | |
3 | 378.413 | 153.496 | 31.270 | 67.517 | 17.837 | 0.903 | |
4 | 505.196 | 174.238 | 47.259 | 91.853 | 18.218 | 0.860 |
Tab. 2 Quantitative evaluation of the scale transformation algorithms (Scale: 0.6 m)表2 IKONOS原始影像与尺度转换后影像定量评价表(尺度: 0.6 m) |
方法 | 波段 | 评价指标 | |||||
---|---|---|---|---|---|---|---|
Mean | STD | SD | RMSE | ERGAS | CC | ||
原始影像 | 1 | 457.501 | 88.939 | - | - | - | - |
2 | 499.650 | 134.548 | - | - | - | - | |
3 | 378.310 | 153.395 | - | - | - | - | |
4 | 504.025 | 174.153 | - | - | - | - | |
商业(软件)算法 | 1 | 457.627 | 89.206 | 18.418 | 42.521 | 9.291 | 0.893 |
2 | 499.810 | 134.880 | 29.469 | 64.073 | 12.820 | 0.886 | |
3 | 378.526 | 153.714 | 33.565 | 71.888 | 18.991 | 0.890 | |
4 | 504.407 | 174.265 | 49.750 | 100.117 | 19.848 | 0.834 | |
本文算法 | 1 | 457.501 | 88.939 | 0.000 | 0.000 | 0.000 | 1.000 |
2 | 499.650 | 134.548 | 0.000 | 0.000 | 0.000 | 1.000 | |
3 | 378.310 | 153.395 | 0.000 | 0.000 | 0.000 | 1.000 | |
4 | 504.025 | 174.153 | 0.000 | 0.000 | 0.000 | 1.000 | |
商业(软件)算法→本文算法 | 1 | 457.683 | 89.222 | 15.082 | 35.732 | 7.807 | 0.925 |
2 | 499.734 | 134.973 | 23.810 | 53.983 | 10.803 | 0.920 | |
3 | 378.419 | 153.820 | 27.105 | 59.563 | 15.740 | 0.916 | |
4 | 504.108 | 174.187 | 39.576 | 79.526 | 15.776 | 0.896 | |
本文算法→ 商业(软件)算法 | 1 | 457.600 | 89.072 | 17.133 | 40.822 | 8.921 | 0.899 |
2 | 499.773 | 134.750 | 27.604 | 61.405 | 12.286 | 0.898 | |
3 | 378.509 | 153.654 | 31.560 | 69.278 | 18.303 | 0.897 | |
4 | 504.139 | 174.248 | 46.784 | 97.203 | 19.281 | 0.844 |
Tab. 3 Quantitative evaluation of the scale transformation algorithms (Scale: 1.5 m)表3 IKONOS原始影像与尺度转换后影像定量评价表(尺度: 1.5 m) |
方法 | 波段 | 评价指标 | |||||
---|---|---|---|---|---|---|---|
Mean | STD | SD | RMSE | ERGAS | CC | ||
原始影像 | 1 | 457.501 | 88.939 | - | - | - | - |
2 | 499.650 | 134.548 | - | - | - | - | |
3 | 378.310 | 153.395 | - | - | - | - | |
4 | 504.025 | 174.153 | - | - | - | - | |
商业(软件)算法 | 1 | 457.537 | 88.939 | 19.808 | 43.908 | 9.597 | 0.893 |
2 | 499.698 | 134.478 | 31.727 | 66.310 | 13.270 | 0.878 | |
3 | 378.434 | 153.370 | 36.363 | 74.707 | 19.741 | 0.882 | |
4 | 504.384 | 174.331 | 53.269 | 103.042 | 20.429 | 0.826 | |
本文算法 | 1 | 457.501 | 88.939 | 0.000 | 0.000 | 0.000 | 1.000 |
2 | 499.650 | 134.548 | 0.000 | 0.000 | 0.000 | 1.000 | |
3 | 378.310 | 153.395 | 0.000 | 0.000 | 0.000 | 1.000 | |
4 | 504.025 | 174.153 | 0.000 | 0.000 | 0.000 | 1.000 | |
商业(软件)算法→本文算法 | 1 | 457.370 | 88.746 | 17.231 | 40.888 | 8.940 | 0.898 |
2 | 499.360 | 134.352 | 27.665 | 61.670 | 12.350 | 0.895 | |
3 | 378.085 | 153.354 | 31.828 | 69.785 | 18.458 | 0.890 | |
4 | 503.460 | 174.258 | 46.045 | 95.918 | 19.052 | 0.849 | |
本文算法→商业(软件)算法 | 1 | 457.456 | 88.861 | 18.961 | 42.756 | 9.348 | 0.901 |
2 | 499.597 | 134.493 | 30.226 | 64.901 | 12.996 | 0.882 | |
3 | 378.419 | 153.359 | 34.431 | 71.901 | 19.015 | 0.889 | |
4 | 504.299 | 174.313 | 52.045 | 97.358 | 19.338 | 0.843 |
Tab. 4 Quantitative evaluation of the scale transformation algorithms (Scale: 3 m)表4 IKONOS原始影像与尺度转换后影像定量评价表(尺度: 3 m) |
方法 | 波段 | 评价指标 | |||||
---|---|---|---|---|---|---|---|
Mean | STD | SD | RMSE | ERGAS | CC | ||
原始影像 | 1 | 457.501 | 88.939 | - | - | - | - |
2 | 499.650 | 134.548 | - | - | - | - | |
3 | 378.310 | 153.395 | - | - | - | - | |
4 | 504.025 | 174.153 | - | - | - | - | |
商业(软件)算法 | 1 | 457.370 | 88.746 | 17.311 | 40.973 | 8.955 | 0.899 |
2 | 499.360 | 134.352 | 27.686 | 61.452 | 12.297 | 0.896 | |
3 | 378.085 | 153.354 | 31.668 | 69.343 | 18.320 | 0.891 | |
4 | 503.460 | 174.258 | 46.766 | 97.201 | 19.279 | 0.845 | |
本文算法 | 1 | 457.501 | 88.939 | 0.000 | 0.000 | 0.000 | 1.000 |
2 | 499.650 | 134.548 | 0.000 | 0.000 | 0.000 | 1.000 | |
3 | 378.310 | 153.395 | 0.000 | 0.000 | 0.000 | 1.000 | |
4 | 504.025 | 174.153 | 0.000 | 0.000 | 0.000 | 1.000 | |
商业(软件)算法→本文算法 | 1 | 457.468 | 89.035 | 11.533 | 34.388 | 7.517 | 0.935 |
2 | 499.554 | 134.604 | 18.368 | 51.286 | 10.266 | 0.929 | |
3 | 378.162 | 153.450 | 21.182 | 58.116 | 15.368 | 0.910 | |
4 | 504.145 | 174.560 | 29.918 | 77.752 | 15.423 | 0.901 | |
本文算法→商业(软件)算法 | 1 | 457.392 | 88.787 | 17.091 | 40.755 | 8.910 | 0.900 |
2 | 499.400 | 134.412 | 27.427 | 61.422 | 12.299 | 0.896 | |
3 | 378.131 | 153.421 | 31.565 | 69.502 | 18.381 | 0.897 | |
4 | 503.449 | 174.265 | 45.571 | 95.478 | 18.965 | 0.850 |
The authors have declared that no competing interests exist.
[1] |
|
[2] |
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
|
[9] |
|
[10] |
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
[19] |
|
/
〈 | 〉 |