Method for generating a focused image of an object
View Patent ↗A method is provided for computing a composite image representing a focused image of an object in an application of machine vision in an optical inspection system. An image tessellated into focus regions is evaluated by region for fine feature sharpness. A sharpness image is computed for each focused region using a fine feature sharpness measurement. A focused composite image is computed by combining as a weighted average, the images at several focus settings, using the sharpness image at each focus setting as the weight. The focused composite image can be further analyzed, inspected, or otherwise processed.
1. A method for generating a focused image of an object from an optical imaging system, the method comprising:
providing a plurality of images of the object, each image having a focus setting;
locating the object in at least one of the plurality of images to provide a known location of the object;
defining a plurality of image regions using the known location of the object;
measuring a sharpness score for each image region of at least two of the plurality of images;
determining a spatial weighting for the image regions using the sharpness score; and
computing a composite image of the object by combining each of the plurality of images using the spatial weighting.
2. The method of claim 1 wherein the step of defining a plurality of image regions further comprises:
determining a set of focus regions on the surface of the object; and
aligning at least one focus region in at least one of the plurality of images.
3. The method of claim 1 wherein the at least one of the plurality of image regions overlaps an adjacent image region using a fuzzy transition.
4. The method of claim 3 wherein the fuzzy transition is a function employing one of the set comprising sigmoid, gaussian and linear.
5. The method of claim 2 wherein the set of focus regions have a fuzzy transition.
6. The method of claim 5 wherein the fuzzy transition is a function employing one of the set comprising sigmoid, gaussian and linear.
7. The method of claim 1 wherein the plurality of image regions comprises a greyscale image map.
8. The method of claim 1 wherein the step of providing a plurality of images further comprises:
determining a coarse focus position.
9. The method of claim 1 wherein the step of providing a plurality of images further comprises:
determining a coarse focus position; and
acquiring a plurality of images at an incremental focus setting.
10. The method of claim 2 wherein the object is a fiber optic cable end face.
11. The method of claim 10 wherein the set of regions are annular.
12. The method of claim 1 wherein the step of measuring a sharpness score further comprises:
transforming each of the image regions of the at least two of the plurality of images so as to provide a plurality of spatial frequencies of the image regions;
measuring a density of high spatial frequencies; and
using the density of high spatial frequencies so as to provide a sharpness score.
13. The method of claim 1 wherein at least one of the plurality of image regions corresponding to a location on the object is defined such that the entire region will be in focus in at least one focus setting.