- Timestamp:
- 08/01/12 17:19:51 (12 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
trunk/python/edgemarker.py
r2613 r2616 18 18 complicated observed area. Typical case is that the area has 19 19 clear 'dent' (e.g. a composite area consisting of two diamond- 20 shaped areas that partially overlap). In such case, the class20 shaped areas that slightly overlap). In such case, the class 21 21 will fail to detect such feature. 22 22 … … 64 64 """ 65 65 Set options for edge detection. Valid options depend on 66 whether scan pattern is raster or not. 66 whether scan pattern is raster or not (i.e. constructor 67 is called with israster=True or False). 67 68 68 FOR RASTER:69 === for raster (israster=True) === 69 70 fraction -- Fraction of OFF integration in each raster 70 71 row. Either numerical value (<1.0) or string … … 86 87 specification by npts will come before. 87 88 88 FOR NON-RASTER:89 === for non-raster (israster=False) === 89 90 fraction -- Fraction of edge area with respect to whole 90 91 observed area. Either numerical value (<1.0) … … 93 94 is same as 0.1. Default is 0.1. 94 95 width -- Pixel width for edge detection. It should be given 95 as a fraction with respect to the median spatial 96 separation between neighboring integrations in time. 97 Default is 0.5. In the most case, default value 98 will be fine. Larger value will cause worse result. 99 Smaller value may improve result. However, if too 100 small value is set (e.g. 1.0e-5), the algorithm may 101 not work. 96 as a fraction of the median spatial separation 97 between neighboring integrations in time. Default 98 is 0.5. In the most case, default value will be fine. 99 Larger value will cause worse result. Smaller value 100 may improve result. However, if too small value is 101 set (e.g. 1.0e-5), the algorithm may not work. 102 102 elongated -- Set True only if observed area is extremely 103 103 elongated in one direction. Default is False.
Note:
See TracChangeset
for help on using the changeset viewer.