The present invention generally relates to image sensors, and more particularly to a tunable color image sensor architecture having floating gate devices.
Presently, semiconductor based image sensors are widely used, especially in digital cameras. These cameras exist as both stand-alone devices and as integrated in other multipurpose electronic devices such as cell phones or personal digital assistants (PDAs).
The marketplace continues to drive for improvements in image capture. These improvements include not only the quality of the captured images, such as improved color accuracy and resolution, but also the lowest possible cost and smallest size.
A common method to capture color information in cameras using semiconductor based image sensors is to employee a mosaic, such as a Bayer pattern, of alternating red, green, and blue pixels. The light reaching these pixels is filtered by corresponding red, green, or blue light filter films made out of materials such as polyimide. However, this method causes several problems. First, conventional RGB filters significantly attenuate the light. It is common for the green and red filters to have less than 40% transmittance and the blue filter to have less than 25% transmittance. This reduction in transmitted light leads to a major decrease in the sensitivity of the sensor array. Second, the film filters commonly used have considerable spectrum overlap. For example, the red filter allows some green light to pass, the green filter allows some red and blue light to pass, and the blue filter allows some green light and red light to pass. This makes determination of the true colors of an image very difficult. Third, the general use of a mosaic of pixels, such as a Bayer pattern, to determine color, degrades the potential spatial resolution of a given array. The amount of degradation depends on lighting conditions but can approach 75% for high resolution systems. For example, a lens with a module transfer function (MTF) value of 0.7 at 100 lp/mm may only provide an MTF of 0.7 at 25 lp/mm for a module with a Bayer mosaic semiconductor image sensor.
Another known method for capturing color information involves Foveon sensors. With the Foveon design, color filtering is achieved by selective absorption of the light through the sensor semiconductor substrate. Like the systems using color films, the Foveon type system still suffers from considerable color spectrum overlap between the different color sensors, thus making the determination of true color by the system difficult. For example, the signal from a blue sensor will have about 35% green light contribution and 10% red light contribution under sunlight conditions. This problem is further aggravated by the fact that the contribution from each color will change under different lighting conditions such as office light vs. sunlight.
Another known method for color filtering that could be applied to imagers is disclosed in U.S. Patent Application Publication 2005/0099373 A1. This method uses a conventional passive color comprised of circularly polarizing selective reflection bands of at least four cholesteric filters together with three liquid crystal switches and related retarder layers. Because this filter only works with polarized light, it is not suitable for cameras used in outdoor and most indoor photography as sunlight and most artificial light sources generate non-polarized light.
Accordingly, it is desirable to provide a new color image sensor system. This system employs a tunable stack of liquid crystal display (LCD) filters which are used to pass a sequence of different colors of light to a sensor array. The system also employs a sensor array optimized to store sequential light information. Furthermore, other desirable features and characteristics of the present disclosure will become apparent from the subsequent detailed description of the disclosure and the appended claims, taken in conjunction with the accompanying drawings and this background.
The present disclosure will hereinafter be described in conjunction with the following drawing figures, wherein like numerals denote like elements, and
The following detailed description is merely exemplary in nature and is not intended to limit the invention or the application and uses of the disclosure. Furthermore, there is no intention to be bound by any theory presented in the preceding background or the following detailed description.
Electromagnetic radiation 26 from an image source is directed through a lens 28 and array overlay 30 onto photosensitive components of the individual pixel circuits 24. Two different embodiments of the array overlay 30 are described later. In both cases, the array is used to provide a temporal sequence of varying light filters to the light that is passed to the sensor array 22. Also, in both cases, cholesteric liquid crystal filters can be utilized. Cholesteric liquid crystal filters can be capable of both better light transmission and much more accurate color filtering than the more commonly used film type color filters.
A row selection circuit 32 is used to activate the readout of the pixel circuits 24 in a given row of the image array 22. The output signals from the pixel circuits 24 in the activated row are provided to a column read circuit 34. Column read circuit 34 may be constructed in any number of different manners. For example, column read circuit 34 may comprise a single correlated double sampling (CDS) circuit that selectively reads individual columns of the array 22 when a single row of the array is selected through the row selection circuit 32. In an alternate exemplary embodiment, a plurality of CDS circuits may be used so that each column of the array 22 (or even fewer than all columns) may be concurrently read by a respective CDS circuit. In other exemplary embodiments, circuits providing a single readout from each pixel circuit 24 during a single read cycle may be employed thereby negating the need for CDS circuitry. Preferably, the analog signals from the pixel circuits 24 are converted by the column read circuit 34 to a digital format which is then arranged into an image frame by a frame grabber 36. Timing for the various operations executed by system 20 is preferably coordinated by a clock and timing generator circuit 38 or the like. Frame grabber 36 may itself execute a number of image processing routines (i.e., image compression, enhancement, etc.) or provide image data at output 40 processing by one or more further systems.
Referring to
A cholesteric liquid crystal 56 is disposed between the first and second walls 52, 54. The cholesteric liquid crystal 56 is preferably a dual frequency liquid crystal material that has a dielectric anisotropy of a first sign over a first range of frequencies, and a dielectric anisotropy of a second sign over a second range of frequencies. Dielectric anisotropy is the difference between the relative permittivity measured parallel to the axes of elongated molecules that make up the cholesteric liquid crystal, and the relative permittivity measured perpendicular to the axes of the molecules.
An alignment layer 58 is supported on the first wall 52 facing the cholesteric liquid crystal 56. The alignment layer 58 preferably comprises a rubbed polymer, e.g., polyimide, and causes the cholesteric liquid crystal 56 to align itself with its helical axis 60 perpendicular to the alignment layer 58 and the first wall 52.
First and second electrodes 62, 64 are supported in spaced relation on the second wall 54. First and second electrodes 62, 64 are substantially co-planer as shown, but each may comprise a plurality of interdigitized fingers as shown in U.S. Pat. No. 6,630,982.
A second alignment layer 66 could be used on the second wall 54 facing the cholesteric liquid crystal 56 and covers the first and second electrodes 62, 64. This second alignment layer 66 is preferably a polymer, and more preferably polyamide, which promotes the above mentioned alignment of the cholesteric liquid crystal 56 in which the helical axis 60 is perpendicular to the alignment layer 58 and the first wall 52.
A drive signal source 70 is coupled to the first and second electrodes 62, 64 when a switch 72 is closed. When the switch is open (as shown), the cholesteric liquid crystal 56 is arranged in a known helical structure which typifies cholesteric liquid crystals. As the signal from the drive signal source 70 is changed, the molecules of the cholesteric liquid crystal 56 rotate according to the electrical field direction and this molecular rotation results a reflected wavelength shift from shorter wavelength to longer wavelength. In terms of color, this corresponds to a color shift from blue to red in the visible range. The intensity control is also described in the U.S. Pat. No. 6,630,982. It can be accomplished with dual frequency driving schemes or time division method.
Therefore, the array overlay 30 alternatively may comprise first thru fourth cholesteric liquid crystal filters 48, 49, 50, 51 which are preferably similar in structure. Each of the filters 48, 49, 50, 51 comprise parallel first and second walls 52, 54. The first wall 52 and the second wall 54 are transparent so as to allow light that is modulated to enter the filter 48, 49, 50, 51. The adjoining walls of filters 48, 49, 50, 51 may comprise a single wall.
Referring to
For the exemplary embodiment shown, pixel circuit 24 comprises first, second, and third floating gate semiconductor devices 80, 82, 84, a photosensitive semiconductor device 86 and a pixel control circuit 88. Though floating gate semiconductor devices are shown, it should be understood that any type of transistor could be used. The floating gate semiconductor device 80 includes a drain 90, a source 92, a control gate 94 and a floating gate 96. The floating gate semiconductor device 82 includes a drain 100, a source 102, a control gate 104 and a floating gate 106. The floating gate semiconductor device 84 includes a drain 110, a source 112, a control gate 114 and a floating gate 116. In the illustrated exemplary embodiment, the photosensitive semiconductor device 86 may be a pinned photodiode that is positioned for exposure to electromagnetic radiation from an image that is to be detected. The photodiode 86 of the illustrated exemplary embodiment includes an anode 118 and a cathode 120.
The pixel control circuit 88 is connected to direct the floating gates of semiconductor devices 80, 82, 84 and the photodiode 86 to a plurality of controlled modes. These controlled modes include at least an erase mode and an exposure mode, and may include a read mode and a data retention mode. In the erase mode, at least a portion of an electric charge is removed from the floating gates 96, 106, 116. The voltage across photodiode 86 may also be raised while in the erase mode. In this manner, both the floating gate semiconductor devices 80, 82, 84 and photodiode 86 are placed in an initialized state.
In the exposure mode, the floating gates 96, 106, 116 are charged at least partially in response to a voltage at a terminal of the photosensitive semiconductor device 86. In the illustrated exemplary embodiment, the floating gates 96, 106, 116 are charged at least partially in response to the voltage at the anode 118 of photodiode 86. The voltage at anode 118 is dependent on the degree to which photodiode 86 is exposed to the electromagnetic radiation from the image source. More particularly, there will be a voltage drop across photodiode 86 that corresponds to the electromagnetic radiation exposure. The greater the exposure that photodiode 86 experiences, the greater the voltage drop that will occur across photodiode 86 thereby reducing the voltage at control gates 94, 104, 114.
Pixel control circuit 88 may also direct photodiode 86 and floating gate semiconductor devices 80, 82, 84 to a data retention mode. In the data retention mode, the charge on the floating gates 96, 106, 116 acquired during the exposure mode is maintained. Notably, the charge on the floating gates 96, 106, 116 remain generally constant even though the voltage drop across photodiode 86 may change. For example, once the floating gates 96, 106, 116 have been charged during the exposure mode, the charge may be maintained on the floating gates 96, 106, 116 almost indefinitely even if the photodiode 86 continues to be exposed to electromagnetic radiation from the image source.
Pixel control circuit 88 may also direct photodiode 86 and floating gate semiconductor devices 80, 82, 84 to a read mode to effectively sense the charge placed on floating gates 96, 106, 116 during the exposure mode. In the illustrated exemplary embodiment, the charge on floating gates 96, 106, 116 alters the threshold voltage VT of the floating gate semiconductor devices 80, 82, 84. Consequently, a predetermined voltage VCS may be provided between the control gates 94, 104, 114 and sources 92, 102, 112, respectively, to produce a current 122 between the drain 90 and source 92 that corresponds to the charge on floating gate 96, a current 124 between the drain 100 and source 102 that corresponds to the charge on floating gate 106, and a current 126 between the drain 110 and source 112 that corresponds to the charge on floating gate 116.
As shown, pixel control circuit 88 may include a transistor switch 130 and a diode 132. Transistor switch 130 may be a field effect transistor, such as a MOSFET or the like, having a drain 134, source 136 and control gate 138. Control gate 138 is connected to receive a row read signal from, for example, row selection circuit 32 of
The various modes of operation mentioned above will now be explained. Exemplary voltage levels for operating in these modes are identified. However, it will be recognized that the specific voltage levels required to operate the pixel architecture 22 in the various modes will depend on the characteristics of the individual devices that are employed.
In the erase mode of operation, drains 134, 122, 124, and 126 as well as sources 92, 102, 112 are driven to +8 V while the row read signal at gate 138 and the reset/erase signal at cathode 144 are driven to −8 V. This places floating gate semiconductors 80, 82, 84 and MOSFET 130 into non-conductive states so that currents 122, 124, 126 and current 146 are approximately zero. The diode 132 is forward biased to discharge floating gates 96, 106, 116. At least a portion of the resulting discharge current is directed to the anode 140. Additionally, photodiode 86 is charged to an initial state with a voltage drop of approximately 15.2 VDC.
In the exposure mode of operation, drains 90, 100, 110 and cathode 144 are sequentially driven to +8 V in synchronization with the tunable color filter while the row read signal at gate 138 and sources 92, 102, 112 are driven to 0 V. This places MOSFET 130 and diode 132 into non-conductive states so that current 146 and current 148 are approximately zero. Additionally, the voltage levels at drain 134 and cathode 120 are elevated to a “programming voltage” of +12 V. Photodiode 86 is exposed to electromagnetic radiation 26 which causes a corresponding decrease in the initial state voltage between the cathode 120 and anode 118. The voltage at control gates 94, 104, 114 reflects this voltage drop and thus corresponds to the amount of electromagnetic radiation detected at photodiode 86. This control gate voltage, in turn, determines the amount of charge placed on floating gates 96, 106, 116 during the exposure mode.
In the data retention mode of operation cathode 144 is driven to +8 V while the row read signal at gate 138 and sources 92, 102, 112 are driven to 0 V. This places MOSFET 130 and diode 132 into non-conductive states so that current 146 and current 148 are approximately zero. The voltage level at cathode 120 of photodiode 86 is reduced to +8 V, thereby inhibiting further accumulation of charge on the floating gates 96, 106, 116. Drains 90, 100, 110 are open circuited, or otherwise connected to a high impedance load to prevent current flow through the floating gate semiconductor devices 80, 82, 84. Currents 122, 124, 126 are therefore approximately zero. In this state, the charge on floating gates 96, 106, 116 can remain relatively constant over a prolonged period of time. Since the charge on floating gates 96, 106, 116 can be retained within the individual pixel circuits 24 of the image array 22, the image processing requirements imposed on peripheral circuits, if any, can be relaxed. The cost and complexity of any such image processing peripheral circuits can therefore be reduced, if desired.
In the read mode of operation, drains 134, 90, 100, and 110, gate 138 and cathode 144 are driven to +8 V while the sources 92, 102, 112 are driven to 0 V. This places control gates 94, 104, 114 at a fixed voltage of approximately +8 V with respect to sources 92, 102, 112, respectively. As such, VCS is approximately +8 V and the currents 122, 124, 126 proceeding selectively from the pixel select circuit 46 through the pixel output corresponds to the charge on floating gates 96, 106, 116. Conversion of currents 122, 124, 126 into an appropriate digital signal may take place in the column read circuit 34, which may be implemented in any number of different manners as understood by those of ordinary skill in the art.
The sequence of steps to capture image sensor information, including the control of both the sensor array and the array overlay, as illustrated in
Red=White−Cyan
Green=Cyan+Yellow−White
Blue=White−Yellow
Other combinations of filters are acceptable. One such example would be: 1) no filter so the sensor sees white light, 2) a blue filter so the sensor sees yellow light, and 3) a green filter so the sensor sees purple light. The information obtained with this combination of filter settings can also be used to derive RGB (Red Green Blue) information as follows:
Red=Purple+Yellow−White
Green=White−Purple
Blue=White−Yellow
The steps are as follows:
The pixel architecture 24 is easily implemented in a monolithic substrate. More particularly, the pixel architecture 24 may be readily manufactured using existing CMOS manufacturing processes to form the image array 22 shown in
Because pixel architecture 24 is centered about floating gate semiconductor devices 80, 82, 84, the pixel, including the components necessary to implement the global reset function, can be implemented with fewer components when compared to a commonly used five transistor pixel architecture. In the specific pixel circuit architecture 24 shown in
The reduction in the number of components employed to implement the pixel circuit 24 can be used to achieve any number of different objectives. For example, pixel circuit 24 may be implemented with a smaller substrate area than a five transistor design of comparable resolution and light sensitivity. A smaller substrate should be lower in cost to manufacture and may allow for a smaller overall imaging system. Alternatively, the pixel design could be implemented using the same size substrate as a five transistor design while offering larger photodiodes. Large photodiodes could allow for improved low light sensitivity and dynamic range. As disclosed herein, the pixel circuit 24 may employ higher operating voltages during the exposure mode, thereby improving the performance of photodiode 86.
Pixel circuit 24 may also be implemented so that the read mode of operation is similar to the readout methods employed in conventional CMOS image sensors. For example, each pixel circuit 24 may be individually addressed to achieve the same windowing and sub-sampling advantages that exist in conventional CMOS sensors thereby obviating the need for substantial redesign of corresponding peripheral readout components. Further, the floating gate semiconductor devices 80, 82, 84 do not have charge leakage issue and they does not have charge recombination issues as a result of under visible light illumination. Thus, it does not have the fading issues associated with the commonly used five transistor CMOS architectures.
One embodiment of a cellular phone 152 that may include a camera that employs the image acquisition system 20 is shown in
While at least one exemplary embodiment has been presented in the foregoing detailed description, it should be appreciated that a vast number of variations exist. It should also be appreciated that the exemplary embodiment or exemplary embodiments are only examples, and are not intended to limit the scope, applicability, or configuration of the disclosure in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient road map for implementing exemplary embodiments of the disclosure, it being understood that various changes may be made in the function and arrangement of elements described in an exemplary embodiment without departing from the scope of the disclosure as set forth in the appended claims.
Number | Name | Date | Kind |
---|---|---|---|
6545653 | Takahara et al. | Apr 2003 | B1 |
6630982 | Li | Oct 2003 | B2 |
6864557 | Turner et al. | Mar 2005 | B2 |
7129466 | Iwasaki | Oct 2006 | B2 |
7508434 | He et al. | Mar 2009 | B2 |
20020171102 | Shizukuishi | Nov 2002 | A1 |
20030098918 | Miller | May 2003 | A1 |
20040178466 | Merrill et al. | Sep 2004 | A1 |
20050062906 | Ishizaki | Mar 2005 | A1 |
20050099373 | Funfschilling et al. | May 2005 | A1 |
Number | Date | Country |
---|---|---|
9819455 | May 1998 | WO |
0005874 | Feb 2000 | WO |
Number | Date | Country | |
---|---|---|---|
20070046794 A1 | Mar 2007 | US |