Television – Camera – system and detail – Solid-state image sensor
Reexamination Certificate
2001-04-18
2004-12-14
Christensen, Andrew (Department: 2615)
Television
Camera, system and detail
Solid-state image sensor
C348S294000, C348S362000, C348S222100
Reexamination Certificate
active
06831689
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the field of optical imagers. More particularly, the invention relates to a method for adaptively expanding the dynamic range of optical active pixel sensors, by using on-chip, real-time automatic scaling of each pixel.
BACKGROUND OF THE INVENTION
Optical imagers (sensors) are widely used in many imaging applications, such as metrology, avionics and space, and particularly as light sensing devices in electronic cameras. Charge Coupled Devices (CCDs) are widely used in optical imagers, and have 65 to 75 dB dynamic range. However, CCDs require special processing, and also lack itegrability. An emerging technology is currently exploited for manufacturing Metal Oxide Semiconductor (CMOS) Active Pixel Sensors (APS), which consume less power and are potentially of lower cost and high integrability.
A typical Active Pixel Sensor (APS) consists of an array of active pixels, each active pixel containing a light sensing element (e.g., a photo-diode or a photo-transistor) and one or more transistors to buffer and amplify the electric signals generated by the light sensing element, in response to light exposure.
Basically, optical imagers may be exposed to a wide range of illumination levels when imaging scenes. Night vision usually introduces illumination levels starting from 10
−3
lux, whereas indoor lighting ranges between 10
2
and 10
3
lux, and sunlight may reach 10
5
lux. This wide range requires a wide dynamic range from the sensors employed.
The dynamic range of a pixel is defined as 20*log(S/N), where S and N are the saturation level of the pixel, and the noise floor, respectively. Low dynamic range entails saturation of pixels with high sensitivity, in case of high illumination levels, or high noise levels, of pixels with lower sensitivity. In both cases, part of the information is lost. Typical dynamic ranges of an APS also range between 65 to 75 dB.
An important factor affecting the dynamic range is the pixel integration time (the time lapses between reset and sample signal), during which the pixel is exposed to illumination and outputs a corresponding electric signal. Basically, a relatively long integration time is required for low illumination levels, so as to obtain a signal which is well beyond the noise floor. On the other hand, a relatively short integration time is required for high illumination levels, so as to eliminate saturation.
Several known methods are used for widening the dynamic range of an APS, which fall into three basic categories: compressing the pixel response curve, clipping the pixel response curve and controlling the pixel integration time. The first two are less advantageous, since they result in loss of information. The latter is preferred, and can be done either globally or locally.
Israeli Patent 100620 describes a method for increasing the dynamic range of optical sensors by conditionally applying a chain of reset signals, within the frame time. A control circuit, which may be common to a group of pixels, compares the illumination levels to a threshold level, which indicates impending saturation, and enables a reset signal after the threshold has been reached. The number of resets during the frame time is counted, and used to calculate a scaling factor, by which the output electric signal is multiplied. However, the circuitry which is required to carry out the comparisons at different time points, which is described in this patent, occupies relatively large area. Hence, the fill-factor (which is the ratio between the pixel area which is responsive to light, and the total pixel area) of the imager is reduced, causing a deterioration in the resolution of the imager.
WO 97/17800 describes an imaging device, comprising an array of pixels having two sampling capacitor banks. Each row is sampled and copied to each capacitor bank twice, first for the short integration time and second, for the long integration time, thereby widening the dynamic range. However, this method is advantageous only if the actual illumination level matches one of these integration times. Any illumination level that falls in between, results in a loss of information. Furthermore, storing the outputs for additional integration times requires more memory cells, which occupy more space, thereby reducing the Field Of View (FOV).
All the methods described above have not yet provided satisfactory solutions to the task of expanding the dynamic range of optical imagers, in real time, and without losing information.
It is an object of the present invention to provide a method and apparatus for expanding the dynamic range of optical imagers, which overcome the drawbacks of prior art imagers.
It is another object of the invention to provide a method for expanding the dynamic range of optical imagers, in real time and during the frame time.
It is still another object of the invention to provide a method for expanding the dynamic range of optical imagers, without losing temporal resolution of the imager, and with minimal effect on spatial resolution.
It is yet another object of the invention to provide a method for expanding the dynamic range of optical imagers, that matches any illumination level.
Other purposes and advantages of the invention will appear as the description proceeds.
SUMMARY OF THE INVENTION
The invention is directed to a method for expanding the dynamic range of an optical imager by individually controlling the integration time of each pixel of a sensor array, and providing a corresponding scaling factor for the electrical output of each individual pixel during the frame time. The integration time of each pixel is controlled as a function of light intensity received thereon, by resetting the pixel after a predetermined threshold for the output signal has been reached. The imager is constructed from a two dimensional active pixel array of M (integer) columns and N (integer) rows, fabricated on a semiconductor substrate. Each individual pixel contains an optical sensor to receive light, a reset input and an electrical output representing the illumination received thereon. The outputs of a selected row are copied into an upper column-parallel signal chain of M capacitors and compared to a set of corresponding threshold values, and into a lower column-parallel signal chain of M capacitors for readout. Preferably, the electrical readout of each pixel is output as an analog signal, or converted to a digital representation. The comparison results are stored in a digital memory. A control circuit controls the reading operations of each pixel, the timing of comparisons for each pixel, and provides corresponding reset signals for each pixel. The required expansion of the dynamic range is determined by a series of W bits, and comparisons are carried out at W time points, having prefixed intervals. Preferably, these time points are selected according to the downgoing series
T
-
T
U
1
,
T
-
T
U
2
,
…
⁢
,
T
-
T
U
W
,
wherein U>1 and T represents the full integration time. Time and space are multiplexed by matching between each time point and a row which is shifted from a selected row n, which is selected for comparison, by a prefixed number of rows. Preferably, row shifts are selected according to the integer values of the downgoing series
n
-
N
U
1
,
n
-
N
U
2
,
…
⁢
,
n
-
N
U
W
.
Preferably, U is selected to be 2, so as to simplify the matching between
T
-
T
2
1
,
T
-
T
2
2
,
…
⁢
,
T
-
T
2
W
and
n
-
N
2
1
,
n
-
N
2
2
,
…
⁢
,
n
-
N
2
W
,
required for time-space multiplexing. Preferably, the downgoing series of row shifts may be also selected as the integer values of the series
N
X
1
,
N
X
2
,
…
⁢
,
N
X
W
where X
i
>1 and i=1, 2, . . . , W. Preferably, the downgoing series of time points may be
T
-
T
X
1
,
T
-
T
X
2
,
…
⁢
,
T
-
T
X
W
.
The comparison results are stored in a memory, and a reset pulse is applied for those pixels that are expected to be saturated, only if that pixel was reset in the precedi
Christensen Andrew
Genco Brian
Lerner David Littenberg Krumholz & Mentlik LLP
Orly Yadid-Pecht
LandOfFree
Optical imager using a method for adaptive real-time... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Optical imager using a method for adaptive real-time..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Optical imager using a method for adaptive real-time... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3278830