Interleaved mosaic imaging rain sensor

Information

  • Patent Grant
  • 6573490
  • Patent Number
    6,573,490
  • Date Filed
    Thursday, June 28, 2001
    25 years ago
  • Date Issued
    Tuesday, June 3, 2003
    23 years ago
Abstract
A plurality of the first L1 and second L2 image sensor elements are interleaved with one another. The first image sensor elements L1 has a peak response at the first wavelength λ1 for capturing an illuminated image of the glass and objects on the glass containing a first wavelength λ1 from an infrared illuminator (12). The second image sensor element L2 has a peak response at a second wavelength λ2 for capturing an ambient image of light rays at the second wavelength λ2 passing through the glass (10) from the exterior to the interior surfaces simultaneously with the capture by the first image sensor element L1. Instead of storing two complete sequential images and differentially subtracting them, the present invention creates two simultaneous images from a specially interleaved, single imager (14).
Description




BACKGROUND OF THE INVENTION




1. Field of the Invention




The subject invention relates to a rain sensor for detecting moisture on the windshield of a vehicle for turning on the windshield wipers, moisture taking the form of mist, water drops, as a film, frost, ice, snow, fog, etc.




2. Description of the Prior Art




The co-pending application Ser. No. 09/835,755 filed Apr. 16, 2001 Attorney Docket No.: 65165.007 entitled DIFFERENTIAL IMAGING RAIN SENSOR discloses a method for achieving greatly improved moisture detection performance through the application of synchronous, differential detection principles. In essence, acquisition and storage of a baseline signal value allows that signal level to be subtracted from a subsequent measurement that includes this background value plus a level indicative of the measured variable. In accordance therewith, that invention provides a method for sensing moisture on the exterior surface of a sheet of glass including the step of capturing an ambient image of light rays passing through the glass from the exterior to the interior surfaces thereof under ambient light conditions. The method is characterized by illuminating the glass with light rays, capturing an illuminated image of the glass and objects on the glass containing the ambient image plus reflected light from the illuminating of the glass, and subtracting the ambient image from the illuminated image to produce a moisture signal.




Ambient light and spurious (optical) signals contribute the main interfering signals in optical rain sensing. Characterizing and storing these signals permits their subsequent effective subtraction.




As disclosed in the co-pending application, a pulsed infrared (IR) light source is used to alternately illuminate image frames for a differential imaging rain sensor. The I.R. illuminant is used to obtain a signal proportional to the density of water drops that collect on the outer surface of a windshield. The use of synchronous detection permits a high degree of selectivity to be achieved since the radiation reflected from the water droplets is unique and can be discriminated over the background light level. Extraneous interfering signals may be subtracted differentially to obtain a net droplet reflection signal and thus a measure of water droplet density. Implicit in this method however, is the requirement that the background-subtracted signal remain invariant during the acquisition of the two images. Naturally, if the reference signal (frame) changes in the interval between successive image capture the basic operating premise is no longer valid. A stable reference requirement implies rapid image acquisition, storage and processing.




Alternately illuminating the area of interest on the windshield surface at high rates also requires that a high intensity pulsed radiant IR source be capable of high frequency modulation. Only solid state (light emitting diode-LED) and pulsed discharge sources are suitable for this purpose, as low cost incandescent lamps cannot be pulsed rapidly enough to be useful.




Differential, alternate image operation of an imaging rain sensor does significantly improve raindrop selectivity at moderate sensitivity. However the technique can be subject to poor false trigger immunity in certain conditions. The net imaging rate for a differential system is at least half that of a conventional (single ended) approach, and from experimental evidence, more like three times slower than the gross image capture rate. The reason for this disparity is that at least two image frames are necessary to yield one net differential frame; and in actual practice a third, wasted interleaving frame is often necessary as well.




Fast framing, typically at sixty net frames per second, is needed to prevent undesired activation (mis-triggering) of the wiping system by spurious, i.e., non-rain, signals. These non-rain artifacts are commonly image components that are not correlated in both differential image frames. Such signals are typically due to moving light sources or pulsed lights that are imaged onto the focal plane array. Given the need for rapid image frame acquisition and processing to prevent false triggering by extraneous lights, the use of differential image processing does in fact reduce rain sensor selectivity under certain conditions. In other words, trying to maintain a high degree of water droplet sensitivity will cause the system to mis-trigger or phantom wipe because it cannot acquire and process images quickly enough to reject spurious artifacts.




In spite of the application of other signal processing enhancements such as narrow band optical filtering and low f/number optics, the selectivity of prior-art imaging rain-sensors needs to be improved for many operating conditions. Certainly the advent of faster cameras and more rapid image processing will improve this situation in the future, but there is a compelling reason to use the current generation of very low cost integrated imagers, even with their lower framing rates.




SUMMARY OF THE INVENTION AND ADVANTAGES




In accordance with the subject invention, moisture on the exterior surface of a sheet of glass is sensed by illuminating the glass and objects on the glass with light rays at a first wavelength, capturing an illuminated image of the glass containing the first wavelength, simultaneously capturing an ambient image of light rays at a second wavelength passing through the glass from the exterior to the interior surfaces thereof, and comparing the images to produce a moisture signal.




This is accomplished with an assembly comprising an illuminator for illuminating the glass and objects on the glass with light rays at a first wavelength, a first image sensor element having a peak response at the first wavelength for capturing an illuminated image of the glass containing the first wavelength, a second image sensor element having a peak response at a second wavelength for capturing an ambient image of light rays at the second wavelength passing through the glass from the exterior to the interior surfaces thereof simultaneously with the capture by the first image sensor element, and a processor for comparing the simultaneously captured images to produce a moisture signal.




The present invention therefore provides an improvement wherein the imager is capable of capturing high rate, single frames that are then processed to yield a differential signal used to activate a wiper system.











BRIEF DESCRIPTION OF THE DRAWINGS




Other advantages of the present invention will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings wherein:





FIG. 1

is a block diagram of an assembly employing the subject invention;





FIG. 2

is a graph showing the first and second wavelengths;





FIG. 3

is a fragmentary view showing a column or row array of first and second image sensor elements;





FIG. 4

is a fragmentary view showing a checkerboard array of first and second image sensor elements; and





FIG. 5

is graphic presentation of the sensing steps to produce the moisture signal.











DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT




Referring to the Figures, wherein like numerals indicate like or corresponding parts throughout the several views, an assembly for sensing moisture on the exterior surface of a sheet of glass


10


is illustrated in FIG.


1


. The assembly comprises an illuminator


12


for illuminating the glass and objects on the glass with light rays at a first wavelength λ1, preferably a continuously illuminated infrared light.




Also included is a first image sensor element L1 having a peak response at the first wavelength λ1 for capturing an illuminated image of the glass containing the first wavelength λ1. A second image sensor element L2 has a peak response at a second wavelength λ2 for capturing an ambient image of light rays at the second wavelength λ2 passing through the glass


10


from the exterior to the interior surfaces thereof simultaneously with the capture by said first image sensor element L1. Actually, the assembly includes a plurality of the first image sensor elements L1 and a plurality of the second image sensor elements L2, the first plurality L1 being interleaved with the second plurality L2 of image sensor elements. The interleaving may be by the first plurality L1 being in alternate horizontal rows or vertical columns, as shown in

FIG. 3

, or in a checkerboard array, as shown in FIG.


4


. Each sensor element comprises an individual pixel disposed in an array over the single planer array


14


.




Instead of storing two complete sequential images and differentially subtracting them, the present invention creates two simultaneous images from a specially interleaved, single imager


14


. Similar image sensor elements are used in color video cameras to create the primary red, green and blue signals that are then combined to produce a full color image. Normally, a set of mosaic or stripe red, green and blue color filters, on the surface of the focal plane imager is used to perform the color separation. Instead of the traditional tri-color filters, the present invention would utilize a mosaic or stripe array of infrared band pass filters as shown in

FIGS. 3 and 4

. One filter over the first sensor elements L1 is chosen so that it selectively and preferentially transmits the first wavelengths λ1 associated with the IR illuminator


12


, while the second filter over the second sensor elements L2 is picked to selectively and preferentially transmit a nearby wavelength, not radiated to any degree by the illuminator


12


. Typically the spacing of these first λ1 and second λ2 wavelengths would be on the order of 100 nm, and both frequencies would be nominally transmitted by the windshield material (glass


10


and coatings) and contained in similar amounts in ambient light, e.g., daylight.




The difference in spectral response of the first L1 and second L2 sensor elements is designed to impart a differential sensitivity to adjacent pixels in the focal-plane imaging array


14


. That is, adjacent, alternating pixels, i.e., the first L1 and second L2 sensor elements, would exhibit a peak response at wavelengths λ1 or λ2 depending upon their physical position on the imager


14


. The best resolution would be achieved if only single pixels were filtered by the alternating stripes or filter tiles as shown in FIG.


3


. The checkerboard (tile or mosaic) bandpass filters shown in

FIG. 4

may be applied by well known micro printing methods, using organic dyes or inks. Alternatively, photolithographic techniques could be used to create the filter array


14


. Multi-layer interference or dielectric filters may be used to obtain enhanced wavelength selectivity but adequate performance has been seen in more conventional, resin-dye filters. While it may be argued that that such printed filters effectively decimate the imager and reduce resolution, the large number of active pixels in even the most modest imagers


14


allows for such division. Accordingly, a first filter may be disposed over each of one of the first L1 and second L2 image sensor elements for filtering light to one of the first λ1 and second λ2 wavelengths respectively.




A lens


16


is included for focusing the images onto the first L1 and second L2 image sensor elements. An optional infrared filter


18


may be employed to filter the illumination from the illuminator


12


. The luminance signals derived from the selectively filtered pixels (sensor elements L1 and L2) in the focal plane array


14


are a function of the image projected onto the array


14


by the lens


16


or optical system.




As illustrated in

FIG. 1

, the assembly includes a processor for comparing the simultaneously captured images to produce a moisture signal. More specifically, the processor includes a camera control and frame capture unit


20


, an image detector


22


, an image storage unit


24


, and a differential comparator


26


. The comparator


26


algebraically sums the first and second images and an analyzer


28


analyzes and sets preset setpoints for initiating the moisture signal, i.e., the wiper activation signal


30


.




The pixel luminance values of immediately adjacent pixels in a given row or column are subtracted from one another to create a differential signal. This difference signal may then be used to determine the presence of bona fide water drops on the windshield surface. In actual use, all rows or columns would contribute their respective delta or difference values that would then be added or summed algebraically to develop an aggregate difference level for the entire video frame. Such summation of all individual sensor elements L1 and L2 (pixel) differences will result in a more accurate picture of the water droplet density on the windshield


10


surface.




As noted, essentially simultaneous capture of the two image frames (by the unit


20


) is advantageous for image processing purposes. Because the image acquisition rate is substantially increased, image smear is virtually eliminated so that moving objects in the imager field of view may be detected and disregarded. Spurious operation due to mistriggering is thereby minimized.




A diagrammatic representation of pixel luminance values is shown in

FIG. 5

, with alternate adjacent pixels being selectively filtered for either first λ1 or second λ2 wavelength response. That is, alternating pixels will be either most sensitive to ambient light λ2 or to the radiant energy λ1 from the illuminator


12


. One means of deriving useful differential information from a single imager


14


involves the subtraction of signal levels from adjacent pixels that are sensitive to different wavelengths, (λ1 or λ2). The response graph shown for L1 merely indicates the expected pixel by pixel signals in an imager row, when illuminated by an image representing several small raindrops on the surface of the windshield


10


. The assumption being that an infrared illuminating source


12


with peak emission at λ1 is preferentially reflected by drops on the glass


10


. The image of those drops on one line (row or column) of the focal plane array


14


is shown schematically, occupying only the first fifteen pixels in that given row or column. As noted, alternate pixels are filtered differently, with either L1 or L2 filtering, providing either the L1 or the L2 filter bandpass response shown in FIG.


2


.




The resulting subtraction of the adjacent pixel output values, L2n−L1


(n+1)


where n is the pixel number (index) in a given row, is shown as the differential response signal, which is directly proportional to the presence and density of water drops on the imaged surface. Note that not all of the drops are ‘counted’, and those drops with optical signatures smaller than one pixel may, in fact, fall on L1 filtered pixels, which are insensitive to the illumination


12


wavelength. (λ1).




It will be appreciated that a row is normally horizontal and a column is normally vertical, either may be employed and are equivalent or the same, even if mis-oriented.




In the implementation of a checkerboard (mosaic or stripe) filter disclosed in

FIG. 1

, the detected image is dissected to provide alternate luminance values for adjacent pixels. These luminance signals are digitally processed to yield differential values corresponding to the image features projected onto the focal plane array


14


. Either singly (or on a row by row or column by column basis) or in aggregate (on a frame by frame basis) these difference values are processed to effect a wiper activation signal


30


. Naturally, various filtering and/or averaging algorithms could be used in the unit


28


to optimize system response to varying conditions, such as light or heavy rain, fog, snow, etc.




The dual wavelength, intrinsically differential image sensor element of the present invention not only permits high image framing rates, but allows for the use of non-pulsed infrared illumination sources. Pulsed sources are not precluded as long as the source


12


can completely illuminate an entire image frame


14


. Very low cost, efficient infrared (IR) sources such as incandescent lamps could be used to great advantage as these devices exhibit excellent performance and long life when operated at moderate filament temperature. In practice, these lamps would be filtered to attenuate any visible light that might distract a driver.




As will be appreciated, the invention provides a method for sensing moisture on the exterior surface of a sheet of glass


10


comprising the steps of illuminating


12


the glass and objects on the glass with light rays at a first wavelength λ1, capturing


20


an illuminated image of the glass and objects on the glass


10


containing the first wavelength λ1, while simultaneously capturing


20


an ambient image of light rays at a second wavelength λ2 passing through the glass


10


from the exterior to the interior surfaces thereof. Also included is the step of comparing


26


the images to produce a moisture signal


30


.




The capturing


20


is further defined as capturing the illuminated image at a first plurality of image sensor elements L1 and capturing the ambient image at a second plurality of image sensor elements L2 interleaved with the first plurality L1 of image sensor elements. Preferably, the first wavelength λ1 is separated from the second wavelength λ2 by one hundred nm. The first plurality of image sensor elements L1 exhibit peak response at the first wavelength λ1 and the second plurality of image sensor elements L2 exhibit peak performance at the second wavelength λ2.




Obviously, many modifications and variations of the present invention are possible in light of the above teachings. The invention may be practiced otherwise than as specifically described within the scope of the appended claims, wherein that which is prior art is antecedent to the novelty set forth in the “characterized by” clause. The novelty is meant to be particularly and distinctly recited in the “characterized by” clause whereas the antecedent recitations merely set forth the old and well-known combination in which the invention resides. These antecedent recitations should be interpreted to cover any combination in which the incentive novelty exercises its utility. In addition, the reference numerals in the claims are merely for convenience and are not to be read in any way as limiting.



Claims
  • 1. A method for sensing moisture on the exterior surface of a sheet of glass (10) comprising the steps of;illuminating the glass (10) and objects on the glass with light rays at a first wavelength λ1, capturing an illuminated image of the glass (10) containing the first wavelength λ1 at a first plurality of image sensor elements, simultaneously capturing an ambient image of light rays at a second wavelength λ2 passing through the glass (10) from the exterior to the interior surfaces thereof at a second plurality of image sensor elements interleaved with the first plurality of image sensor elements, and comparing the images to produce a moisture signal.
  • 2. A method as set forth in claim 1 wherein the first wavelength λ1 is separated from the second wavelength λ2 by one hundred nm.
  • 3. A method as set forth in claim 1 wherein the first plurality of image sensor elements exhibit peak response at the first wavelength λ1 and the second plurality of image sensor elements exhibit peak performance at the second wavelength λ2.
  • 4. An assembly for sensing moisture on the exterior surface of a sheet of glass (10) comprising;an illuminator (12) for illuminating the glass (10) on objects on the glass with light rays at a first wavelength λ1, a plurality of first image sensor elements L1 having a peak response at the first wavelength λ1 for capturing an illuminated image of the glass (10) containing the first wavelength λ1, a plurality of second image sensor elements L2 interleaved with said first image sensor elements L1 and having a peak response at a second wavelength λ2 for capturing an ambient of light rays at the second wavelength λ2 passing through the glass (10) from the exterior to the interior surfaces thereof simultaneously with the capture by said first image sensor elements L1, and a processor for comparing the simultaneously captured images to produce a moisture signal.
  • 5. An assembly as set forth in claim 4 wherein said first and second plurality of image elements are disposed in one of alternating rows and columns.
  • 6. An assembly as set forth in claim 4 wherein said first and second plurality of image sensor elements are disposed in a checkerboard array.
  • 7. An assembly as set forth in claim 4 wherein each of said sensor elements comprises a pixel.
  • 8. An assembly as set forth in claim 7 including a first filter disposed over each of said first and second image sensor elements L1, L2 for filtering light to one of said first and second wavelqngths λ1, λ2.
  • 9. An assembly as set forth in claim 4 including a lens for focusing said images onto said sensor elements.
  • 10. An assembly as set forth in claim 9 wherein said processsor includes a comparator for algebraically summing the first and second images and an analyzer for setting preset setpoints for initiating the moisture signal.
RELATED APPLICATION

This application is an improvement over co-pending application Ser. No. 09/835,755 filed Apr. 16, 2001, and assigned to the assignee of the subject invention.

US Referenced Citations (24)
Number Name Date Kind
4164868 Suntola Aug 1979 A
4476419 Fukatsu et al. Oct 1984 A
4595866 Fukatsu et al. Jun 1986 A
4636643 Nakamura et al. Jan 1987 A
4639831 Iyoda Jan 1987 A
4676638 Yasuda Jun 1987 A
4805070 Koontz et al. Feb 1989 A
4827198 Mueller et al. May 1989 A
4831493 Wilson et al. May 1989 A
4867561 Fujii et al. Sep 1989 A
4960996 Hochstein Oct 1990 A
5059877 Teder Oct 1991 A
5313072 Vachss May 1994 A
5537003 Bechtel et al. Jul 1996 A
5668478 Buschur Sep 1997 A
RE35762 Zimmerman Apr 1998 E
5923027 Stam et al. Jul 1999 A
6020704 Buschur Feb 2000 A
6084519 Coulling et al. Jul 2000 A
6144022 Tenenbaum et al. Nov 2000 A
6262410 Stam et al. Jul 2001 B1
6320176 Schofield et al. Nov 2001 B1
6376824 Michenfelder et al. Apr 2002 B1
6392218 Kuehnle May 2002 B1
Foreign Referenced Citations (3)
Number Date Country
19742093 Mar 1999 DE
9189533 Jul 1997 JP
WO 9729926 Aug 1997 WO