Image analysis – Image enhancement or restoration – Intensity – brightness – contrast – or shading correction
Reexamination Certificate
1997-01-29
2001-03-27
Bella, Matthew C. (Department: 2621)
Image analysis
Image enhancement or restoration
Intensity, brightness, contrast, or shading correction
C382S270000, C358S451000
Reexamination Certificate
active
06208766
ABSTRACT:
BACKGROUND AND SUMMARY OF THE INVENTION
This invention relates to a process for modifying the quality of images which are furnished by an electronic imaging system in the form of image point signals representing gray-scale values.
A process of this generic type is discussed in
SPIE
, Vol. 238, “Image Processing for Missile Guidance” (1980), Pages 119 to 125. This document concerns images which are furnished by an electronic imaging system, and are composed of a plurality of image points representing gray scale values, which are distributed areally in an orderly manner. Rectangular image segments are defined around each of the individual image points, and from the image point signals f(x,y) of the image points situated within these image segments, chracteristic brightness values H representing the local brightness as well as characteristic contrast values K, representing the local contrast are calculated and assigned to the individual image points. The image point signals f(x,y) are then modified by means of these characteristic values, specifically in the sense of influencing the background brightness and the contrast component. The resulting modified image point signals g(x,y) originate from the original image point signals f(x,y) according to the following rule:
g(x,y)=kH+c[f(x,y)−H]/K
The first of these two terms may be called the background brightness and the second may be called the contrast component of the image point signal. The characteristic brightness value H is calculated in a manner which will be explained later, by taking the mean of all image points existing within the respective image segment. Therefore, it represents a characteristic brightness value which relates to this image segment; that is, which is local and which first acts as the measurement for the background brightness. The individual brightness deviation of the individual image point signals f(x,y) from the assigned characteristic brightness value H, multiplied with the quotient c/K, forms the individual contrast component of the respective image point signals. The background brightness can be varied by means of the constant k; the contrast fraction can be varied by means of the constant c. The characteristic contrast value K is derived by taking, for all image points of the respective image segment, the mean of the deviations of the individual image point signals from the characteristic brightness value, as will be explained in detail in the following. Its introduction into the above-mentioned rule has the result that the individual contrast components of the image point signals are modified inversely proportionally to the characteristic contrast value. The background brightness of the individual image points is levelled according to the background brightness of the local environment and the contrast component is amplified in the case of a slight environmental contrast and is damped in the case of a high environmental contrast.
Therefore, in this image processing method, a rectangular image segment is defined which surrounds each individual image point, all such image segments having the same size. By taking the mean of all image points situated within each such image segment, the characteristic brightness value to be assigned to the respective image segment as well as the characteristic contrast value are calculated. The image point signals are successively subjected to the modification according to the above-mentioned rule, and considerable computing expenditures are required for each individual image point. Moreover, these expenditures increase as larger image segments are selected. Such computing expenditures are not acceptable in every case.
It is therefore an object of the invention to provide a process of the initially mentioned type which minimizes computing expenditures.
This object is achieved by the image processing system according to the invention in which a separate image segment, for which the respective extensive calculations of the characteristic brightness and characteristic contrast value must be carried out, is not assigned to every individual image point within the whole image Rather only selected individual image points are utilized which, in the following, will be called supporting points. The selection is made such that, using a suitable definition of the image segments surrounding the supporting points, at least the areas of the image which are of interest (and generally therefore the whole image) are completely covered by means of these image segments. Therefore, the respective characteristic brightness and contrast values need be calculated only for a number of image segments which corresponds to the number of selected supporting points; these characteristic values are then assigned to the pertaining supporting points. Since the number of image segments necessary to cover the whole image can be orders of magnitude lower than the number of all image points existing in the image, a significant reduction of the computing expenditures is achieved. The characteristic brightness and contrast values determined for the individual image segments can then be used in the above-mentioned sense for the modification of the image point signals of the image points pertaining to the respective image segment.
To modify the image point signals furnished by the electronic imaging system as described above, the above-mentioned characteristic values are required. These may be determined from the incoming image point signals themselves, and also from the already modified image point signals.
In a first of two embodiments of the invention, based on the characteristic values assigned to the supporting points, characteristic values are generated for each of the remaining image points by means of a two-dimensional interpolation. The interpolated characteristic values are, in turn, used to modify the image point signals of the remaining image points. Thus, from the grid of the original totality of image points, a limited number of supporting points are first selected, the characteristic values are calculated only for the supporting points and, from this limited number of characteristic values, by means of a two-dimensional interpolation, a number of characteristic-value pairs are obtained which correspond to the original grid and to the original number of image points. Such interpolation is an arithmetically simple operation which requires only limited computing expenditures (much lower than the computing expenditures necessary in the case of the initially mentioned process), and does not depend on the number and size of the selected image segments. After the interpolation process is completed, similarly to the initially discussed process, an individual pair of characteristic values is assigned to each image point, so that the background brightness and the contrast component (described in detail below) can be individually modified in each case.
According to the second embodiment of the invention, the computed characteristic values are used for the collective modification of all image point signals of the respective image segment of the current image. In this case, to compute the characteristic values to be assigned to the supporting points, it is necessary to use the already modified image point signals of the preceding image.
Therefore, while, in the case of the first embodiment of the invention, modification of the individual image point signals of an image is made possible by interpolation of the characteristic values, the second embodiment is limited to carrying out the modification of the image point signals of the respective image segment in a global manner, in which case, in each new image, the result of the modification of the preceding image is taken into account by way of the characteristic values determined from the latter.
As a further development of the second embodiment, the respective characteristic contrast values are evaluated and the image point signals are modified as a function of the result of this evaluation in different manners. On the one h
Dzida Karl
Schweyer Nikolaus
Bella Matthew C.
Deutsche Aerospace AG
Evenson, McKeown, Edwards & Lenahan P.L.L.C.
LandOfFree
Process for influencing the quality images furnished by an... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for influencing the quality images furnished by an..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for influencing the quality images furnished by an... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2486736