Image analysis – Histogram processing – For setting a threshold
Reexamination Certificate
1998-04-30
2004-03-30
Wu, Jingge (Department: 2623)
Image analysis
Histogram processing
For setting a threshold
C382S169000
Reexamination Certificate
active
06714674
ABSTRACT:
TECHNICAL FIELD
The present invention relates to digital imaging and, more particularly, to the conversion of a range of input pixel values to a different range of output pixel values representative of an image, such as an X-ray image.
BACKGROUND ART
Discrete imaging devices, such as digital X-ray imaging systems, employ a detector which divides regions of an image into individual picture elements, or pixels. The array or matrix of pixels defines, when viewed as an overall image, features of interest, such as internal anatomy of a subject positioned adjacent to the detector. To facilitate interpretation by physicians and technicians, the individual intensities of the pixels typically define the features of the image by imitating contrasts and textures obtainable through conventional film-based X-ray or imaging systems.
To convert the detected pixel intensities to digitized values suitable for display, the pixel intensity values are processed after acquisition by the detector. In a first stage, the detected pixel intensities are digitized in values which vary over a predetermined dynamic range of the detector and acquisition circuitry, such as 12 to 14 bits. In X-ray systems, for example, these digitized values are representative of the quantity of X-rays received by each pixel during data acquisition. Subsequently, the pixel intensity values are scaled to map the values onto the dynamic range of a display device. As part of this scaling, it is common to perform logarithmic transformation of the image pixel values to obtain a resulting image which mimics conventional film-rendered images. In addition, the scaling process maps the dynamic range of the detector and acquisition circuitry onto the dynamic range of the display. The latter range may be substantially different from that of the upstream circuitry, such as on the order of 8 to 10 bits.
While the logarithmic transformation of the digitized pixel values is useful in rendering an image which is understandable by attending physicians and technicians, performing the transformation prior to the dynamic range scaling can be problematic. For example, histograms are often employed to analyze pixel intensity values. However, processing of histograms generated based on the transformed values can result in difficulties in identifying high and low limits of relevant portions of the detected data, rendering the dynamic range scaling difficult. The use of logarithmically transformed data prior to dynamic range scaling can also result in loss of accuracy for individual pixels in the image matrix.
There is a need, therefore, for an improved method for processing discrete image data that facilitates use of as much of the dynamic ranges of acquisition circuitry and display circuitry as possible. In particular, there is a need for an improved method for converting digital pixel values defining a discrete pixel image from a first dynamic range to a second dynamic range in a computationally efficient manner.
DISCLOSURE OF THE INVENTION
In an exemplary embodiment of the invention, digitized values for a plurality of pixels of an image, such as an X-ray image, are converted from a first dynamic range to a second dynamic range for eventual display of the overall image. The input values, distributed over the first dynamic range, are used to form a histogram representative of the number of pixels having predetermined digital intensity values. A lower limit of the relevant portion of the input data dynamic range is identified from the histogram. A transformed histogram is developed based upon the input value histogram. The transformed value histogram is used to determine a threshold value for the log-transformed image. Based upon this threshold value, an upper limit of the relevant portion of the input data range is identified from the transformed value histogram. With the lower and upper limits of the input value dynamic range thus identified, the input pixel values are mapped onto output values over the second dynamic range.
The technique offers a computationally efficient approach to mapping of input and output dynamic ranges in image processing, particularly in digital X-ray systems. By identifying the low and high limits of the input dynamic range, the useful range of the input values is identified and the mapping is performed over this useful range. The technique may be employed for individual images, thereby accommodating variations in intensity values for individual patients and individual images. The resulting images provide consistent appearance, facilitating comparison and interpretation of imaged features.
REFERENCES:
patent: 4950894 (1990-08-01), Hara et al.
patent: 5060081 (1991-10-01), Shimura
patent: 5164993 (1992-11-01), Capozzi et al.
patent: 5239378 (1993-08-01), Tsuji et al.
patent: 5305204 (1994-04-01), Ohhashi
patent: 5633511 (1997-05-01), Lee et al.
patent: 5748773 (1998-05-01), Tashiro et al.
patent: WO89/12278 (1989-12-01), None
Schwenker, et al., Automatic Detection of the Useful Image Data From Digital X-ray Detectors, pp. 204-209, 1997,Society for Photo-Optical Instrumentation Engineering.
Schwenker, et al., Automatic Detection of the Useful Image Data From Digital X-ray Detectors,Society for Photo-Optical Instrumentation Engineering, vol. 3031, pp. 204-209, 1997.
Fletcher Yoder
General Electric Company
Wu Jingge
LandOfFree
Method for converting digital image pixel values does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for converting digital image pixel values, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for converting digital image pixel values will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3265331