Image analysis – Image sensing
Reexamination Certificate
2000-04-07
2003-06-24
Grant, II, Jerome (Department: 2729)
Image analysis
Image sensing
C348S687000
Reexamination Certificate
active
06584238
ABSTRACT:
The invention relates to a dynamic range modification and to a camera comprising such a dynamic range modification. The notion modification includes compression and expansion.
The article “Adaptive highlight compression in today's CCD cameras”, SMPTE Journal, March 1992, pp. 135-139, discloses an adaptive variable knee proposal to make the best use of the increased dynamic range of current CCD sensors, as displays (typically having a maximum dynamic range of 40 dB) cannot deal with the large dynamic range (being over 75 dB) of CCD sensors. The same knee characteristic is active in all three color channels (R, G, B).
The article “The all-digital camcorder—the arrival of electronic cinematography”, SMPTE Journal, January 1996, pp. 13-30, discloses a digital processing circuit having a gamma/knee circuit for each of the three digitized color channels R, G, B. Several circuits, including an A/D converter and a matrix circuit, are located between each of the three CCD sensors (one for each color) and the gamma/knee circuit.
It is, inter alia, an object of the invention to provide a simpler modification, especially one that is suitable for use with a single CCD color camera. To this end, a first aspect of the invention provides a dynamic range modification as defined in claims 1 and 6. A second aspect of the invention provides a camera comprising such a dynamic range modification. Advantageous embodiments are defined in the dependent claims.
In a dynamic range modification in accordance with a primary aspect of the present invention, a brightness component is selected from a sensor output signal, the brightness component is non-linearly processed to provide a non-linearly processed brightness signal, and a modified signal is furnished in dependence upon the non-linearly processed brightness signal.
Preferably, in the sensor output signal, after the sample & hold operation, a non-linear processing operation (compression, expansion or histogram egalization) is carried out on the brightness component, while a linear brightness-level-dependent gain correction is carried out on the color component.
The invention is based on the following recognitions. Especially in security cameras there is a need for rendering details visible under circumstances where the illumination of a scene may cause large contrast problems. Also in consumer cameras there is a need for such a dynamic range modification. While in a 3-CCD camera three brightness signals are available for R, G, and B, respectively, the sensor signal from a camera having only one CCD sensor contains both brightness and color information. A non-linear processing operation on the latter signal would result in intolerable errors in the color information. For this reason, brightness and color information in the sensor output signal are currently first transformed into Y, R, G, B signals, to which Y, R, G, B signals compression is applied. To bring an overload margin of, for example, 600% back within the range of the system, all signal processing circuits preceding the dynamic range compression should be linear up to at least 600%. In general, this demand cannot be met by standard analog processing ICs. For digital ICs, this requirement implies an input accuracy of over 13 bits. This is feasible for studio cameras, but for consumer and security cameras such a high accuracy is very costly and not available in compact form. In most cases, the dynamic range modification will be a dynamic range compression. However, for example in misty conditions, it might be necessary to carry out a dynamic range expansion rather than a dynamic range compression to obtain a good contrast.
The invention is based on the recognition that it is possible to apply a dynamic range modification on the output signal of a CCD sensor having a color filter by reducing high amplitudes only for those signals which primarily include a brightness component. So, the A/D conversion circuit and the matrix circuit receive an already modified signal which renders it possible to use cheaper circuits than if the modification took place after the A/D conversion circuit and the matrix circuit like in the prior art. The color filter may, for example, be a Bayer filter or a complementary mosaic filter.
This separate processing of the brightness and color components in a sensor output signal is based on the recognition that this is possible because these components occupy different places in the frequency domain, as will now be explained. The surface of a CCD sensor can be interpreted as a sensitive layer having a window structure, each window representing a pixel. The pixel information is read with a clock frequency which is high enough to read all pixels on a row within one TV line period. Consequently, the output signal of a CCD sensor has a frequency spectrum in which the brightness information is present at baseband and around all harmonics m*fs of the sample frequency fs. The color information results from a color filter on the sensor surface, alternate pixels being covered by alternate colors. With two alternating colors in the horizontal read-out direction, a repeating frequency spectrum with a frequency offset of half (m*fs/2) the pixel read-out frequency is obtained at the output.
So, if the sensor signal passes a low-pass filter, a representation of the brightness information is available.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.
REFERENCES:
patent: 4926247 (1990-05-01), Nagasaki et al.
patent: 5138458 (1992-08-01), Nagasaki
patent: 5638138 (1997-06-01), Hickman
patent: 5815291 (1998-09-01), Shono et al.
patent: 5963665 (1999-10-01), Kim et al.
patent: 6384937 (2002-05-01), Matama
patent: 404342371 (1992-11-01), None
“Adaptive Highlight Compression in Today's CCD Cameras”, H. Blom et al, SMPTE Journal, Mar. 1992, pp. 135-139.
“The All-Digital Camcorder—The Arrival of Electronic Cinematography”, Laurence J. Thorpe and A. Takeuchi, SMPTE Journal, Jan. 1996, pp. 13-30.
“New Signal-Processing LSIS for the 8MM Camcorder” Makota Onga et al, IEEE Transactions on Consumer Electronics, vol. 36, No. 3, Aug. 1990, pp. 494-501.
“False Color Signal Reduction Method for Single-Chip Color Video Cameras”, Hiroaki Sugiura et al, IEEE Transactions on Consumer Electronics, vol. 40, No. 2, May 1994, pp. 100-106.
“Adaptive Gamma Processing of the Video Cameras for the Expansion of the Dynamic Range”, Sigeo Sakaue et al, IEEE Transactions on Consumer Electronics, vol. 41, No. 3, Aug. 1995, p. 555-561.
Dielhof Pieter B.
Van Der Heijden Ralph M. W.
Belk Michael E.
Grant II Jerome
Koninklijke Philips Electronics , N.V.
LandOfFree
Dynamic range modification does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Dynamic range modification, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dynamic range modification will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3134462