Method and apparatus for digital image defect correction and noise filtering

A method for detecting and correcting detector aberration defects in digital diagnostic images, the method comprising:

• generating diagnostic image data;
• organizing the diagnostic image data into a two or three-dimensional array of image pixel values;
• merging adjacent pixels of like value into a single pixel value to generate grouped pixel values;
• selecting an n×n kernel of pixels to be examined, where n is an integer greater than 2;
• calculating a median pixel value of the groupled pixel values;
• comparing pixel values of the kernel with threshold criteria;
• (1) in response to the compared pixel value passing the threshold criteria, retaining the compared pixel value,
• (2) in response to the compared pixel value failing the threshold criteria, replacing the pixel value with the median pixel value;
• whereby, the retained and replaced pixel values form a corrected digital diagnostic image.

A method for detecting and correcting defects in digital diagnostic images comprising:

• generating diagnostic image data;
• organizing the diagnostic image data into a two or three-dimensional array of pixels;
• selecting an n×n kernel of pixels to be examined, where n is an integer greater than 2;
• merging adjacent pixels of like value into a single pixel value to generate grouped intensity values;
• calculating a median value of the grouped intensity values;
• in response to generating a group of three or more intensity values;
• comparing the center pixel value to the median value,
• determining whether the center pixel value deviates by more than a preselected percentage from the median value, and,
• based on a comparison, one of retaining the center pixel value and replacing the center pixel value with a value calculated from neighboring pixel values;
• and
• in response to generating a group of two intensity values;
• determining whether a first intensity value is one of at or below a preselected intensity value for black pixels and at or above a preselected intensity value for white pixels, and,
• based on a comparison, one of replacing the first intensity value with a second intensity value and replacing the second intensity value with the first intensity value.

In a radiographic apparatus having a penetrating radiation source for projecting x-rays across an examination region, a plurality of radiation detector arrays disposed across the x-ray examination region from the penetrating source, the detector arrays including analog to digital converters for converting analog signals into digital signals, and an image calibration processor for calibrating and reconstructing the digital signals into pixel values of a digital image representation, an adaptive median filter for detecting and correcting defects in the digital image representation due to defective detectors in the radiation detector arrays, the adaptive median filter comprising:

• a memory which stores each candidate pixel value and neighboring pixel values;
• a processor which:
• calculates a reference value from the neighboring pixel values,
• compares a relationship between the candidate pixel value and the reference value with an adaptive threshold criterion which is calculated by multiplying the reference value by a preselected percentage;
• based on the comparison, one of (1) replacing the candidate pixel value with a function of the neighboring pixel values and (2) retaining the candidate pixel value.

A method of radiographic diagnostic imaging including:

• propagating x-rays through a subject;
• detecting the x-rays which have propagated through the subject;
• converting the detected x-rays into light signals;
• converting the light signals into electrical signals;
• generating an image representation;
• scanning the image representation on a pixel by pixel basis using an n×n kernel;
• calculating a median value of each kernel including:
• reordering each kernel in numerical sequence,
• merging adjacent pixels of like value into a merged pixel, and
• after merging, selecting a numerical central pixel value as the median value;
• calculating a deviation between the median value and the scanned pixel value;
• comparing the deviation to a defect threshold;
• and
• based on the comparison, one of (1) replacing the scanned pixel value with the median value and (2) retaining the scanned pixel value.

An adaptive filter for detecting and correcting digital image defects comprising:

• a memory which stores each candidate pixel value and a kernel containing each candidate pixel value and a plurality of neighboring pixel values;
• a processor which reorders the pixel values of each kernel, calculates a median value for the kernel, calculates a difference between the candidate pixel value and the median value, and compares the difference to a defect threshold which is calculated by multiplying the candidate pixel by a preselected percentage;
• and,
• based on the comparison, one of (1) replacing the candidate pixel value with the median value and (2) retaining the candidate pixel value.