This disclosure relates to imaging means and in particular to the enhancement of images from electronic imaging means particularly, but not exclusively, the enhancement of such images which take the form of arrays of rows and columns of pixels and in which the brightness of any pixel or a brightness component of any pixel can be represented numerically, for example, digitally. The disclosure is particularly concerned with enhancement of video images in which a moving scene is represented, in known manner, by a series of briefly presented images or frames.
The disclosure concerns an infrared (IR) vision arrangement for automobiles in which an infrared camera mounted on the front of the vehicle. For example, the vision arrangement may be mounted on the front bumper or behind the windscreen, and views the area in front of the vehicle in infrared light. The vision arrangement produces corresponding video signals which, after processing, are used to drive a display visible to the driver, thus allowing the driver, in conditions of poor visibility in visible light, for example at night and/or in heavy rain or fog, to see the road ahead more clearly, and especially to see warm objects such as pedestrians or animals.
In developing an infrared video system of the kind referred to above for automotive purposes, the applicants have encountered certain problems. One such problem is that the change in overall brightness between successive images will become very large if the warm window of the camera is hit by a splash of water. In one of its aspects, this disclosure provides an image enhancing method, and a corresponding apparatus, to address such problems.
Accordingly, another aspect of the disclosure provides an imaging apparatus including an IR video camera comprising a sensor array that is operable to produce successive video images in the form of pixelated two-dimensional images. The imaging apparatus further comprises a processing arrangement operable to reduce or increase the overall brightness of an image in dependence upon the overall brightness of the pre-enhanced image and the overall brightness of the preceding enhanced image, so that the change in the overall brightness (DC-level) between the enhanced image and the preceding enhanced image is decreased to be less than the difference in overall brightness between the unenhanced image and the preceding enhanced image.
The brightness level of a new image is adjusted by adding, or multiplying by, a term or factor which is calculated from the overall brightness of the pre-enhanced new image and the overall brightness of the preceding image.
Preferably, the output at time t is calculated using
fout(x,y,t)=f(x,y,t)+γ·DCresidual(t) (1)
where
DCresidual(t)=DCout(t)−DCf(t) (2)
DCf(t) being the mean brightness of the input image and
DCout(t)=(1−λ)·DCout(t−1)+λ·DCf(t) (3)
where γ is a parameter controlling the strength of the compensation and λ is a parameter determining the adaptation rate of the compensation.
Another aspect of the present invention provides a image enhancement method applicable to IR video images in which, for each video image, the overall brightness of an enhanced image is reduced or increased in dependence upon the overall brightness of the pre-enhanced image and the overall brightness of the preceding enhanced image.
A further aspect of the present invention provides an image enhancement method applicable to IR video images in which at least a fraction of the overall brightness of successive images from an IR video camera is low-pass filtered with respect to time.
Conveniently, the brightness level of a new image is adjusted by adding, or multiplying by, a term or factor which is calculated from the overall brightness of the pre-enhanced new image and the overall brightness of the preceding image.
Advantageously, the output at time t is calculated using
fout(x,y,t)=f(x,y,t)+γ·DCresidual(t) (1)
where
DCresidual(t)=DCout(t)−DCf(t) (2)
DCf(t) being the mean of the input image and
DCout(t)=(1−λ)·DCout(t−1)+λ·DCf(t) (3)
where γ is a parameter controlling the strength of the compensation and λ is a parameter determining the adaptation rate of the compensation.
Preferably, the overall brightness level of an image is calculated as being, or as being proportional to, the average of the signal levels corresponding to each of the pixels in the frame.
Conveniently, the IR wavelengths observed fall within the far IR portion of the spectrum.
Another aspect of the disclosure provides a vehicle comprising an image enhancement apparatus according to the above.
In order that the disclosure may be more readily understood, embodiments will now be described, by way of example, with reference to the accompanying drawings, in which:
Referring to
In preferred embodiments of the invention the wavelengths observed are in the “far” IR portions of the spectrum, i.e. wavelengths of between around 10 and 1000μ, or more particularly between around 25 and 350μ.
The processor 4 processes the signals, or at least the information in these signals, in the manner described below and provides, to a video display 5, driving signals such that the display 5 presents visibly to the viewer, an enhanced version of the scene viewed in infrared by the camera 1.
With an infrared video system as described above, without the image enhancement technique discussed below, there is a problem, as previously noted, in that, if the camera is mounted in front of the vehicle, and the warm window 3 of the camera 1 is hit by a splash of water, the water will be heated and the average intensity of the image “seen” by the camera 1 will rise suddenly. This problem may be alleviated by removing the “DC component” of the video signal, as mentioned above. One way of achieving this is by using a spatial high pass filter, however, simply removing the DC component would create other distortions of the image.
A better way to compensate for brightness “flashes” arising from water splashes is to modify the DC-level of each new image, by low pass filtering of the overall DC-level from one frame to the next.
For example, the overall brightness of a new image gathered by the camera 1 may be calculated as an average of the signal level of all or some of the pixels in the image. Alternatively, the overall brightness may be considered to be the total brightness of all or some of the image pixels. The raw data f(x,y,t) (where x and y designate horizontal and vertical directions on the array of pixels forming part of the CCD) output by the camera 1 may then be manipulated to give output data fout(x,y,t) using the following formula:
fout(x,y,t)=f(x,y,t)+γ·DCresidual(t) (1)
where
DCresidual(t)=DCout(t)−DCf(t) (2)
DCf(t) being the mean brightness of the input image and
DCout(t)=(1−λ)·DCout(t−1)+λ·DCf(t) (3)
γ is a parameter which controls the strength of the flash compensation. If this parameter is set to 1, the DC-level will be DCout. The adaptation rate of the flash compensation is determined by the parameter λ.
Equation (1) could also be written as:
fout(x,y,t)=f(x,y,t)+γ·(1−λ)·(DCout(t−1)−DCf(t))
In other words, the “raw” brightness of a new image to be displayed by the display is adjusted by adding a term (γ·DCresidual(t)) which is dependent on the overall brightness of the unenhanced new image and of the previous (enhanced) image. DCresidual(t) is calculated as being the difference between the overall brightness of the new (unenhanced) image and a further variable (DCout(t)), which is the sum of a portion of this variable for the previous enhanced frame and a portion of the overall brightness of the unenhanced new frame. The sizes of these portions are preferably determined by multiplying the overall brightnesses by first and second respective coefficients. In preferred embodiments these coefficients add up to 1, and are equal to (1−λ) and λ respectively, where λ is the adaptation rate parameter as mentioned above.
Referring to
In alternative embodiments, the “raw” overall brightness of a new image may be multiplied or divided by an appropriate factor, to adjust the overall brightness in the same way, and a skilled person will appreciate how this may be achieved.
Whichever technique is used, the effect is preferably to adjust the overall brightness of the new frame to be in between its “raw” overall brightness and the overall brightness of the previous, enhanced frame.
It will be appreciated that low pass temporal filtering of at least a fraction of the DC-level of the output images makes the output robust against sudden histogram shifts, for instance rain flashes, as discussed above.
When used in this specification and claims, the terms “comprises” and “comprising” and variations thereof mean that the specified features, steps or integers are included. The terms are not to be interpreted to exclude the presence of other features, steps or components.
The features disclosed in the foregoing description, or the following claims, or the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for attaining the disclosed result, as appropriate, may, separately, or in any combination of such features, be utilised for realising the invention in diverse forms thereof.
Number | Date | Country | Kind |
---|---|---|---|
0522594.1 | Nov 2005 | GB | national |
0603255.1 | Feb 2006 | GB | national |
This application is a continuation of International Patent Application PCT/SE2006/001212 filed Oct. 27, 2006, and thereby claims the benefit of GB 0603255.1 filed Feb. 17, 2006 and GB 0522594.1 filed Nov. 4, 2005.
Number | Name | Date | Kind |
---|---|---|---|
4718028 | Gussin et al. | Jan 1988 | A |
4737847 | Araki et al. | Apr 1988 | A |
5341219 | Schelen et al. | Aug 1994 | A |
5629988 | Burt et al. | May 1997 | A |
7340164 | Charlet et al. | Mar 2008 | B2 |
20030067546 | Asano | Apr 2003 | A1 |
20030095080 | Colmenarez et al. | May 2003 | A1 |
20030142850 | Eggers et al. | Jul 2003 | A1 |
20040066458 | Kawamura et al. | Apr 2004 | A1 |
20040136605 | Seger et al. | Jul 2004 | A1 |
20060188246 | Terre et al. | Aug 2006 | A1 |
Number | Date | Country |
---|---|---|
100 16 184 | Oct 2001 | DE |
0 539 897 | Oct 1992 | EP |
0 542 091 | Nov 1992 | EP |
2 303 757 | Feb 1997 | GB |
1 383 083 | Jul 2003 | GB |
2 388 988 | Nov 2003 | GB |
11-205636 | Jul 1999 | JP |
2005274301 | Oct 2005 | JP |
WO 9525322 | Sep 1995 | WO |
WO 2007053075 | May 2007 | WO |
Number | Date | Country | |
---|---|---|---|
20080204571 A1 | Aug 2008 | US |
Number | Date | Country | |
---|---|---|---|
Parent | PCT/SE2006/001212 | Oct 2006 | US |
Child | 12114542 | US |