Image analysis – Image enhancement or restoration – Intensity – brightness – contrast – or shading correction
Reexamination Certificate
1998-02-24
2001-05-29
Rogers, Scott (Department: 2624)
Image analysis
Image enhancement or restoration
Intensity, brightness, contrast, or shading correction
C382S105000, C382S205000
Reexamination Certificate
active
06240217
ABSTRACT:
This application is related to applications filed at the same time by the same applicant entitled “Imaging Apparatus” (claiming priority from Australian Provisional Application PO5257 filed Feb. 24, 1997) and “Vehicle Imaging and Verification” (claiming priority from Australian Provisional Application PO5256 field Feb. 24, 1997). The contents of those applications are hereby incorporated herein by reference.
This invention relates to a method of processing digital image information. It relates particularly but not exclusively to a method of selectively modifying pixel intensity information in order to improve legibility or visibility of parts of a digital image.
Image contrast problems resulting in under-exposure and over-exposure of parts of an image are relatively common in photography. When an image of a person or object is taken facing into the sun, or with some other background which is bright compared to the person or object, the person's or object's features often are difficult to discern in the finished photographic image. Similarly, when part of the image is very bright compared to the rest of the image, the bright part is often over-exposed or “burnt out”, so that its details are not discernible in the finished photographic image.
This problem is particularly noticeable in consumer video cameras which have automatic light contrast controls based on average light intensity for an image. In order for such video cameras to be able to capture images with a range of different intensities, they increase or decrease the brightness of each image captured, depending on the average brightness of all pixels in the image. When an image is being taken, for example, of a person in an unlit room against the background of a window, with bright sunshine outside, the person will often appear merely as a silhouette against the window, because the camera has compensated for the bright background by reducing the pixel image intensity. If the person moves against a wall in the same room, so that the window is no longer in the background, the average pixel intensity reduces, and the camera compensates by increasing pixel intensities, so that the person's features become visible.
In some video cameras, a manual image intensity override control is provided to allow an operator to compensate for this problem. The operator can move the control to ensure that the image pixel intensities remain bright even when the background is bright. This results in “image burn” (or over-exposure) for the background, but at least it ensures that the details of interest, such as the person's face, remain clearly visible. Alternatively, if the operator is taking images of, say, a fireworks display on a dark night, the camera will naturally tend to brighten the pixel intensities to compensate for the darkness of the average pixel intensity, with the result that the fireworks images are “burnt out”. By using the manual exposure control, the operator can ensure that the pixel intensities remain dark and the fireworks remain visible.
In film-based photography, it is possible to compensate for a bright background or a dark background by varying the length of exposure during development of the film. However, if it is desired that very dark and very bright features both be visible in a photograph, it is often necessary to use two different exposures and create two separate images.
A particular application in which image contrast presents a problem is traffic cameras used in Law Enforcement Programs such as speed and red-light violation prosecution. Due to the retro-reflective nature of most licence plates, any light energy which is directed towards this area is reflected back at a level (result of a high reflection efficiency) which is higher then the average intensity entering the camera. Consequently an optical burn (ie over exposure) appears around the area of the licence plate. An example of this is shown in FIG.
14
.
SUMMARY OF THE INVENTION
An object of the present invention is to provide a method of selectively varying the image intensity of part of an image.
According to a first aspect of the present invention there is provided a method of processing digital image information which includes pixel intensity information, including the steps of:
identifying a region of the image in which pixel intensities are similar to each other, but the median pixel intensity differs significantly from the median pixel intensity of other parts of the image; and
modifying pixel intensities in the identified region so that the median for the region is closer to the median for the other parts of the image.
The identified region of the image may have a median pixel intensity below the median pixel intensity of the rest of the image, as in the previously described example of a person photographed in a dark room against the background of a bright window. The region may comprise the person's face, and the pixels in the region of the person's face may be modified by increasing their intensity values until their median intensity value approaches the median for the rest of the image. This results in an image which shows both the background and the person in good detail.
Alternatively, the identified region may have a median pixel intensity higher than the median for the rest of the image, as in the previously described example of the retro-reflective licence plate. In this case, the identified region is the licence plate, and the pixel intensities in that region are modified by reducing their intensity values until their median intensity value approaches the median for the rest of the image. Again, this results in an image which shows both the background and the licence plate in good detail.
The region may be identified in any suitable way. One way of identifying the region is for a person to examine the image and manually identify any region which needs pixel intensity correction. Another way is locate computationally any pixels which have an intensity greater than a particular high value or less than a particular low value and which have more than a predetermined number of other pixels meeting the same criteria within a predetermined distance. Another way arises when all images being processed have common features—in this case, a predefined region at the same place on each image can be used, if it is known in advance that the region contains pixels which require pixel intensity correction. Alternatively, if it is known that a similar-shaped region requiring pixel intensity correction will appear on all images, but not necessarily in the same place (as in the case of traffic images featuring retro-reflective licence plates), a shape recognition algorithm may be applied to the image to locate the region.
Any suitable pixel intensity modification algorithm may be applied to the identified region. One suitable modification algorithm involves adding a fixed intensity value to, or subtracting a fixed intensity value from, all pixel intensity values in the identified region. Another algorithm involves multiplying or dividing all pixel intensity values in the region by a fixed amount. It will be appreciated that many different algorithms using fixed or sliding scales may be used, depending upon the particular application.
According to a second aspect of the invention, there is provided a method of processing digital image information by modifying pixel intensities according to a non-linear algorithm, with the result that one or more regions of the image have their median pixel intensities adjusted to a level which is closer to the median pixel intensity for other parts of the image.
In this aspect of the invention, the pixel modification algorithm can be applied to the whole of the image, and not just to a particular identified region. For example, the algorithm may specify that all pixel intensity values over a particular threshold are to be reduced according to a particular scale, and all values under the threshold are to be kept constant, or increased or reduced according to another scale. In the case of the
Ercan Gurchan
Whyte Peter
Dergosits & Noah
Redflex Traffic Systems Pty LTD
Rogers Scott
LandOfFree
Digital image processing does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Digital image processing, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Digital image processing will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2486407