Teachable object contour mapping for biology image region partition
View Patent ↗A teachable object contour mapping method for region partition receives an object boundary and a teaching image. An object contour mapping recipe creation is performed using the object boundary and the teaching image to generate object contour mapping recipe output. An object contour mapping is applied to an application image using the object contour mapping recipe and the application image to generate object contour map output. An object region partition using the object contour map to generate object region partition output. An updateable object contour mapping method receives a contour mapping recipe and a validation image. An object contour mapping is performed using the object contour mapping recipe and the validation image to generate validation contour map output. An object region partition receives a region mask to generate validation object region partition output. A boundary correction is performed using the validation object region partition to generate corrected object boundary output. An update contour mapping is performed using the corrected object boundary, the validation image and the contour mapping recipe to generate updated contour mapping recipe output.
1. A teachable object contour mapping method for biological object aggregation comprising the steps of:
a) Input an object boundary and a teaching image containing object aggregate;
b) Perform object contour reference map creation using the object boundary having object contour reference map output wherein the object contour reference map has lowest values at boundary and highest values around center; Perform object center region creation using the object boundary having object center region output; Perform reference mapping using the object center region and the object boundary having the object contour reference map output wherein the object contour reference map is created by combining outer distance map, inner distance map, and height in a ratio format;
c) Perform object contour mapping recipe creation using the object contour reference map and the teaching image having object contour mapping recipe output wherein the object contour mapping recipe stores fitting coefficients for fitting pixels in the teaching image to pixels in the contour reference map;
d) Input an application image;
e) Perform object contour mapping using the object contour mapping recipe and the application image having object contour map output, wherein the object contour mapping applies the fitting coefficients in the object contour mapping recipe to pixels in the application image.
2. The teachable object contour mapping method of claim 1 further comprises an object region partition step using the object contour map having object region partition output wherein the object region partition is performed by a guided partition method.
3. The teachable object contour mapping method of claim 1 wherein the object contour mapping recipe creation step further
performs soft fitting using the object contour reference map and the teaching image having object contour mapping recipe output.
4. The soft fitting step of claim 3 comprises the steps of:
a) Perform feature generation using the teaching image having image features output;
b) Perform feature selection using the image features and the object contour reference map having selected features output;
c) Perform fitting using the selected features and the object contour reference map having object contour mapping recipe output.
5. The object contour mapping recipe creation step of claim 3 uses multi-resolution soft fitting.
6. The object contour mapping recipe creation step of claim 3 inputs different weights for different points of mapping.
7. The reference mapping step of claim 1 combines outer distance map of the object center region and inner distance map within the object boundary by the following rule
Map_reference[ x][y ]=height*Dist_inner_object[ x][y ]/(Dist_inner_object[ x][y ]+Dist_outer_center[ x][y ]).
8. The teachable object contour mapping method of claim 1 wherein the object contour mapping is performed in multi-resolution fashion from lowest resolution level to highest resolution level.