초록 ▼

A method for filtering clutter from a first digital input image is disclosed. Generally, the method includes the steps of applying at least first and second transforms to first digital image data corresponding to the first digital input image, performing an appropriate filtering operation to retain,

A method for filtering clutter from a first digital input image is disclosed. Generally, the method includes the steps of applying at least first and second transforms to first digital image data corresponding to the first digital input image, performing an appropriate filtering operation to retain, in first and second transform domains, at least a portion of the transform indications corresponding or at least correlatable to a structure or target in the first digital input image, applying to the remaining transform indications inverse first and second transforms, and processing the filtered first digital input data to produce a single filtered image and/or image data, whereby the target or structure in the first digital input image formerly obscured by the clutter is detectable/recognizable.

대표청구항 ▼

1. A method for reducing clutter in a first digital image, the first digital image corresponding to first digital image data and including at least a first candidate object of interest obscured by the clutter, wherein the first digital image data is in an image intensity domain, said method comprisi

1. A method for reducing clutter in a first digital image, the first digital image corresponding to first digital image data and including at least a first candidate object of interest obscured by the clutter, wherein the first digital image data is in an image intensity domain, said method comprising the steps of:performing first and second transforms to transform the first digital image data in the image intensity domain into first and second transformed image data in first and second transform domains, respectively, wherein first and second transform indications of each of the first and second transformed image data corresponding to at least the first candidate object of interest and the clutter, respectively, in the first and second transform domains are separable; filtering at least a portion of the second transform indications corresponding to the clutter in the first and second transform domains; performing inverse first and second transforms on at least the first transform indications in the first and second transform domains, respectively, to transform at least the first transform indications into first filtered indications in the image intensity domain, the first filtered indications corresponding to at least the first candidate object of interest; and performing one of first and second steps, said first step comprising the step of combining the first filtered indications in the image intensity domain to produce a first filtered image, said second step comprising processing the first filtered indications in the image intensity domain using a principal component analysis.