Regiongrowing using mean gray values.
regiongrowing_mean performs a regiongrowing using mean gray values of a region, starting from points given by StartRows and StartColumns. At any point in the process the mean gray value of the current region is calculated. Gray values at the boundary of the region are added to the region if they differ from the current mean by less than Tolerance. Regions smaller than MinSize are suppressed.
If no starting points are given (empty tuples), the expansion process starts at the upper leftmost point, and is continued with the first unprocessed point after a region has been created.
|
Image (input_object) |
image(-array) -> object : byte / int4 |
| Image to be segmented. | |
|
Regions (output_object) |
region-array -> object |
| Extracted segments. | |
|
StartRows (input_control) |
point.y(-array) -> integer |
| Row coordinates of the starting points. | |
| Default value: '[]' | |
| Typical range of values: 0 <= StartRows <= 511 (lin) | |
| Minimum increment: 1 | |
|
Recommended increment: 1 | |
|
StartColumns (input_control) |
point.x(-array) -> integer |
| Column coordinates of the starting points. | |
| Default value: '[]' | |
| Typical range of values: 0 <= StartColumns <= 511 (lin) | |
| Minimum increment: 1 | |
|
Recommended increment: 1 | |
|
Tolerance (input_control) |
number -> real |
| Maximum deviation from the mean. | |
| Default value: 5.0 | |
| Suggested values: 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 12.0, 15.0, 17.0, 20.0, 25.0, 30.0, 40.0 | |
| Typical range of values: 0.1 <= Tolerance <= 100.0 (lin) | |
| Minimum increment: 0.01 | |
|
Recommended increment: 1.0 | |
| Restriction: Tolerance > 0.0 | |
|
MinSize (input_control) |
integer -> integer |
| Minimum size of a region. | |
| Default value: 100 | |
| Suggested values: 0, 10, 30, 50, 100, 500, 1000, 2000 | |
| Typical range of values: 0 <= MinSize <= 50000 (lin) | |
| Minimum increment: 1 | |
|
Recommended increment: 100 | |
| Restriction: MinSize >= 0 | |
regiongrowing_mean returns 2 (H_MSG_TRUE) if all parameters are correct. The behavior with respect to the input images and output regions can be determined by setting the values of the flags 'no_object_result', 'empty_region_result', and 'store_empty_region' with set_system. If necessary, an exception is raised.
regiongrowing_mean is reentrant and automatically parallelized (on tuple level).
gauss_image, sigma_image, anisotrope_diff, median_image
select_shape, reduce_domain, opening, expand_region
regiongrowing, regiongrowing_n
Region processing
