Method for non-referential defect characterization using fractal encoding and active contours
A method for identification of anomalous structures, such as defects, includes the steps of providing a digital image and applying fractal encoding to identify a location of at least one anomalous portion of the image. The method does not require a reference image to identify the location of the anomalous portion. The method can further include the step of initializing an active contour based on the location information obtained from the fractal encoding step and deforming an active contour to enhance the boundary delineation of the anomalous portion.
1. A method for identification of anomalous structures including defects, comprising the steps of:
providing a digital image;
applying fractal encoding to identify a location of at least one anomalous portion within said image, and
initializing an active contour based on said location obtained from said fractal encoding step and deforming an active contour to enhance a boundary of said anomalous portion, wherein said image is acquired at a single instant of time, said fractal encoding comprising non-contractive mapping.
2. The method of claim 1 , wherein said active contour is initialized at a perimeter of said location.
3. The method of claim 1 , wherein said fractal encoding comprises mapping a plurality of domain subregions into a plurality of range subregions, further comprising the step of compiling a similarity parameter for each of said range subregions, said similarity parameter based on the ability of each of said range subregions to be represented by said mapping.
4. The method of claim 3 , further comprising the step identifying respective ones of said range subregions having low relative similarity parameter values as suspected anomalous portions.
5. The method of claim 4 , further comprising the steps of combining adjacent ones of said suspected anomalous portions to form combined anomalous portions and identifying a largest of said combined anomalous portions.
6. The method of claim 1 , wherein said at least one anomalous portion comprises a plurality of anomalous portions, further comprising the step of selecting a subset of said plurality of anomalous portions based on at least one of said location and a size of said anomalous portions.
7. The method of claim 1 , wherein said image is an image of a semiconductor circuit.
8. The method of claim 1 , wherein said image is a scanning electron microscope (SEM) image or an image from an optical microscope.
9. The method of claim 1 , wherein said digital image is of a man-made structure.
10. A system for identification of anomalous structures, comprising:
structure for applying fractal encoding to an image, and
structure for identifying a location corresponding to at least one anomalous portion within said image comprising structure for initializing an active contour based on said location obtained from said fractal encoding and deforming an active contour to enhance a boundary relating to said anomalous portion, said location being obtainable without a reference image, wherein said system employs non-contractive fractal encoding.
11. The system of claim 10 , wherein said system employs region selection based on size and location, followed by active contour initialization and iterative contour adjustment to locate borders of said anomalous portion.
12. The system of claim 10 , wherein said digital image is of a man-made structure.