WO2008099146A1 - Method and apparatus for counting vehicle occupants - Google Patents
Method and apparatus for counting vehicle occupants Download PDFInfo
- Publication number
- WO2008099146A1 WO2008099146A1 PCT/GB2008/000460 GB2008000460W WO2008099146A1 WO 2008099146 A1 WO2008099146 A1 WO 2008099146A1 GB 2008000460 W GB2008000460 W GB 2008000460W WO 2008099146 A1 WO2008099146 A1 WO 2008099146A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- infrared
- vehicle
- interior
- wavelengths
- range
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
- G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
- G06V10/143—Sensing or illuminating at different wavelengths
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Studio Devices (AREA)
Abstract
An imaging apparatus (10) for determining the number of occupants in a vehicle, has a means for illuminating an interior of the vehicle (15) within a first range of infrared wavelengths (23), a first capture means (16) for capturing a first infrared image of the interior (15) for the first range of wavelengths, a means for illuminating the interior of the vehicle (15) within a second range of infrared wavelengths (24), a second capture means (17) for capturing a second infrared image of the interior (15) for the second range of wavelengths, and a processor (21) for analysing the first and second captured infrared images of the interior (15) of the vehicle and operable to determine the number of occupants in the vehicle.
Description
Method and Apparatus for Counting Vehicle Occupants
The present invention is directed to methods and apparatus for determining the number of occupants in a vehicle.
There is a requirement to count the number of occupants of a vehicle in a variety of situations. These range from high occupancy vehicle lanes where penalties can be imposed on vehicles with less than a specified number of occupants, for example less than two occupants, to situations where toll or congestion charging is based on the number of vehicle occupants, (this could include car parks for high occupancy vehicles).
It will of course be appreciated that the above examples are not exhaustive and that there may be numerous other situations, including security requirements, in which it may be desirable to determine the number of occupants in a vehicle.
One method for determining the number of occupants in a vehicle is for an observer to view a vehicle and count the number of occupants on board. However, manual observation is generally regarded as being time consuming and inaccurate. Another way of counting the occupants could be to record standard video images of the interior of the car and then to use image processing techniques to determine the number of occupants. However, it may be difficult to differentiate from video images between real people and dummies.
What is preventing the widespread adoption of a number of congestion limiting schemes is the availability of an automatic method of counting the number of occupants in a vehicle. Attempts have been made to develop such automatic methods and apparatus based on the fact that human skin, regardless of ethnic origin, absorbs in the infra red region of the spectrum around 1550 nm, for reference see WO2005119571.
US6370260 teaches the use of two infrared cameras, one capturing a broadband image between 800 nm and 1400 nm, and the other capturing a broadband image above 1400 nm. The differences between the two images allow detection of human skin. The use of two infra red cameras is expensive, and the experience of the present inventors is that the use of broadband images leads to false positives. The absorption of skin at 1550 nm is due to its water content, so any spherical organic material (e.g. a lettuce or a cabbage) could be mistaken for a human head.
WO2005119571 teaches the use of an infrared camera with a 1550 nm filter to obtain an image with dark areas corresponding to human skin, and a normal video camera with a filter to limit its input to the red end of the visible spectrum (600 - 700 nm). In the red part of the visible spectrum human skin reflects, but green vegetation absorbs. Thus differences between the two images are enhanced by human skin appearing light in the visible image and dark in the infra red one. However, even this approach can lead to confusion in certain lighting conditions. By using ambient visible light it can be prone to errors caused by shadowing, and, of course, it cannot be used in the hours of darkness as it is not possible to illuminate the interior of a vehicle with visible light without risking dazzling the driver.
Aspects and embodiments of the invention are set out in the accompanying claims.
In general terms, one embodiment of a first aspect of the present invention can provide imaging apparatus for determining the number of occupants in a vehicle, comprising: means for illuminating the interior of the vehicle within a first range of infrared wavelengths; a first capture means for capturing a first infrared image of the interior for the first range of wavelengths; means for illuminating the interior of the vehicle within a second range of infrared wavelengths; a second capture means for capturing a second infrared
image of the interior for the second range of wavelengths; and a processor for analysing the first and second captured infrared images of the interior of the vehicle and operable to determine the number of occupants in the vehicle.
In particular, the means for illuminating the interior of the vehicle within a second range of infrared wavelengths allows the apparatus to be used at night, without dazzling the driver of the vehicle.
In an embodiment, the second range of infrared wavelengths is between approximately 800nm and approximately 900nm. Preferably, the means for illuminating the interior within the second range of wavelengths is adapted to illuminate the interior with radiation at approximately 808nm.
In one embodiment, the second capture means is a conventional video capture means, such as a video camera.
The inventors have recognised that the infrared wavelengths in the range between approximately 800nm and 900nm are both reflected by human skin, and can also be captured by a normal video camera. Visible wavelengths, of course, end at around 700nm.
Preferably, the first range of infrared wavelengths is between approximately 1400nm and approximately 1600nm. In an embodiment, the means for illuminating the interior within the first range of wavelengths is adapted to illuminate the interior with radiation at approximately 1550nm.
Suitably, the first and second capture means include narrowband filter devices matching the respective wavelengths of the first and second illuminating devices.
In one embodiment, the processor is operable to perform an AND function on the first and second captured infrared images.
One embodiment of a second aspect of the present invention can provide a method for determining the number of occupants in a vehicle, the method comprising: illuminating an interior of the vehicle within a first range of infrared wavelengths; capturing a first infrared image of the interior for the first range of wavelengths; illuminating the interior of the vehicle within a second range of infrared wavelengths; capturing a second infrared image of the interior for the second range of wavelengths; and analysing the first and second captured infrared images of the interior of the vehicle to determine the number of occupants in the vehicle.
Preferably, the second range of infrared wavelengths is between approximately 800nm and approximately 900nm. Suitably, the method comprises illuminating the interior within the second range of wavelengths with radiation at approximately 808nm.
In an embodiment, the method comprises capturing the second infrared image with a conventional video capture means, such as a video camera.
One embodiment of a third aspect of the present invention can provide imaging apparatus for determining the number of occupants in a vehicle, comprising a means of capturing a first infrared image together with a means of illuminating the interior of the vehicle at the first infrared wavelength, a means of capturing a second infrared image together with a means of illuminating the interior of the vehicle at the second infrared wavelength, an optical arrangement adapted to provide a common view of the interior of the vehicle to both means of infrared image capture to enable the capture of two infrared images at different wavelengths of the common view of the interior of the vehicle, and a processor to analyse the captured infrared images of the common view of the interior of the vehicle and able to determine the number of occupants in the vehicle.
One embodiment of a fourth aspect of the present invention can provide a method for determining the number of occupants in a vehicle, the method comprising: providing a common view of an interior of a vehicle to both means of capturing infrared images; capturing infrared images of the common view of the interior of the vehicle using two means of capturing infrared images respectively; analysing the captured infrared images of the common view of the interior of the vehicle to detect the presence of persons in the captured images; and determining the number of occupants in the vehicle based on the image analysis.
Preferably, the method further comprises illuminating the interior of the vehicle at the first infrared wavelength, and illuminating the interior of the vehicle at the second infrared wavelength.
The present invention overcomes the previously perceived difficulties in a number of ways. For example, firstly it makes use of the fact that a conventional video camera will operate up to 1000 nm. Thus by providing illumination into the car in the infrared region between 800 nm and 900 nm, and using a narrowband filter on the video camera matched to the wavelength of the illuminator, images can be captured in which human skin appears pale. Secondly by using an AND function in the detection software, the system looks for both a pale area in the video camera image and a corresponding dark area in the infrared (1550 nm) image. This provides a much more robust result.
The optical arrangement may include a lens arrangement, and the lens arrangement may comprise a single lens, to ensure that a common view of the interior of the vehicle is captured at both wavelengths.
The means of capturing the first infrared image may be operable to capture an image in the wavelength range between approximately 1400nm
and approximately 1600nm, ideally at approximately 1550nm. The means of capturing the second infrared image means may be operable to capture an image in the wavelength range between approximately 800nm and approximately 900nm. The longer wavelength range is chosen to make use of the fact that human skin absorbs infrared radiation in this wavelength range. The shorter wavelength range is chosen to enable normal video cameras to be used without causing dazzle or distraction to the driver (as would be the case if illumination in the visible range of the spectrum was used). It also corresponds to a region of the spectrum where human skin reflects light, in contrast to the behaviour at the longer wavelength.
Radiation in the longer wavelength range is considered to be 'eye safe', since it is absorbed by the cornea. However, radiation in the shorter wavelength range may not be eye safe. The design of the shorter wavelength illuminating device can therefore include diffusing elements such that no person in the vicinity of the apparatus, especially one not protected by the infra red absorbing layers on modern vehicle windscreens, is at risk of eye damage.
In order to minimise false readings, in one embodiment the apparatus include narrow band filter devices so that the images captured match the wavelengths of the two illuminating devices. In an embodiment, in the longer wavelength range the filter device and illuminating device operate at approximately 1550nm to maximise the effect of absorption of infrared radiation by human skin. In another embodiment, in the shorter wavelength range the filter chosen to match the illuminating device (e.g. 808 nm).
The optical arrangement may include a beam splitter device. The beam splitter device may be located between the lens arrangement and the two infrared image capture means. The beam splitter device may be adapted to direct infrared radiation at different wavelengths from the lens arrangement to the respective means of capturing the infrared images. The beam splitter device may be adapted to direct infrared radiation from the lens arrangement
to the infrared image capture means by transmitting the longer wavelength infrared radiation through the beam splitter device, and reflecting the shorter wavelength infrared radiation. The beam splitter device may be adapted to direct the infrared radiation from the lens arrangement to the shorter wavelength image capture means by reflecting the longer wavelength infrared radiation.
The processor may be operable to analyse the captured infrared images of the common view of the vehicle interior by comparing the respective captured images to detect differences between them. It may be preferable for the processor to perform an 'AND' function on the captured infrared images of the vehicle interior such that a face is detected in each.
The step of determining the presence of persons in the vehicle may comprise analysing the detected dark areas to identify the dark areas which correspond to an exposed human face. The step of determining the presence of persons in the vehicle may comprise counting the identified dark areas corresponding to an exposed human face to determine the number of persons present in the vehicle.
The above aspects and embodiments may be combined to provide further aspects and embodiments of the invention.
An embodiment of the present invention will now be described by way of example only and with reference to the accompanying drawings, in which:-
Fig. 1 is a diagrammatic view of imaging apparatus for determining the number of occupants in a vehicle according to an embodiment of the invention, in use;
Fig. 2 is a detailed view of the imaging apparatus of Fig. 1 ; and
Fig. 3 shows examples of images obtained using the imaging apparatus of Fig. 2 according to an embodiment of the invention.
Referring to Fig. 1 , there is shown generally imaging apparatus 10 for determining the number of occupants in a vehicle 14. The imaging apparatus
10 may be mounted in use on a suitable post 11 at the side of the road, or slung from a gantry or bridge over the road using a suitable mounting arrangement 12. The imaging apparatus 10 is arranged so that its elliptical cone of view and illumination 13 falls on the road and the vehicle passes through it. On doing so, the interior of the vehicle 15 is illuminated at the two wavelengths and the two cameras are able to capture images of the interior.
Referring to Fig. 2, the imaging apparatus 10 comprises an infrared image capture means, for example an infrared camera 16, which is arranged to capture an image at the longer infrared wavelength, and a conventional video image capture means, for example a video camera 17, which is arranged to capture an image at the shorter infrared wavelength.
The imaging apparatus 10 includes an optical arrangement 18 which is adapted to provide a common view 19 of the interior 15 of the vehicle 14 to both cameras 16, 17 to enable the infrared and video cameras 16, 17 to capture longer and shorter wavelength infrared images respectively of the common view 19. For clarity purposes, the interior of the vehicle 15 is represented schematically in Fig. 2 by the cross-hatched area.
The imaging apparatus 10 includes a processor 21 which is operable to analyse the captured shorter and longer wavelength infrared images of the common view 19 of the vehicle interior 15 to determine the number of occupants in the vehicle 14, based on the image analysis.
In more detail, the infrared camera 16 is adapted to capture an infrared image of the common view 19 of the vehicle interior 15 in which the
wavelength of the infrared radiation is approximately 1550nm. In order to ensure that the infrared camera 16 receives infrared radiation at this predetermined wavelength only, the imaging apparatus 10 includes a filter device 22, for example a notch filter or band pass filter, which filters the infrared radiation provided by the optical arrangement 18 to the infrared camera 16.
The imaging apparatus 10 also includes an infrared illumination means, such as a laser 23. The laser 23 provides infrared illumination having a predetermined wavelength of approximately 1550nm to the vehicle interior 15. The provision of radiation at a wavelength of 1550nm using the laser 23 is advantageous since, at this wavelength, there is no risk of retinal damage to the vehicle occupants.
A suitable beam spreading device (not shown), for example a telescope, spreads out the emitted laser beam so that substantially the entire vehicle interior 15 is illuminated by the infrared illumination from laser 23.
An infrared image at the longer wavelength is captured by activating the laser 23 to provide infrared illumination to the vehicle interior 15 and activating the infrared camera 16 to capture an infrared image of the vehicle interior 15.
The resultant captured infrared image will be a greyscale image comprising dark and light areas, defined by individual pixels in the image. The level of greyscale, that is the greyscale value, of each pixel will depend upon the reflectance characteristics of objects present in the common view 19 of the vehicle interior 15.
In particular, the exposed skin of any occupants in the vehicle 14 will appear as dark areas, defined by groups of adjacent dark or black pixels, in the captured infrared image due to the fact that human skin absorbs a substantial amount of infrared radiation at the predetermined wavelength of 1550nm.
Human skin absorbs, rather than reflects, infrared radiation at this wavelength
because it consists of a substantial amount of water, the spectral characteristics of water being such that it absorbs infrared radiation at the predetermined wavelength of 1550nm. It will of course be appreciated that water and, therefore, human skin absorbs infrared radiation over a range of wavelengths. The use of infrared radiation having a predetermined wavelength of 1550nm is, however, advantageous for the reasons discussed above and for further reasons which will be explained later.
Other materials inside the vehicle 14, including hair and artificial materials such as clothing and upholstery, will appear as lighter areas in the captured infrared image since these materials will reflect significantly more of the infrared radiation at the predetermined wavelength of 1550nm than human skin. The individual pixels of the infrared image defining these lighter areas will have different greyscale values according to the different reflectance characteristics of the particular materials.
In one embodiment, the imaging apparatus 10 utilises greyscale values in the range 0 to 255, a greyscale value of 0 corresponding to a black pixel, and a greyscale value of 255 corresponding to a white pixel. It will of course be appreciated by those skilled in the art that the imaging apparatus 10 could utilise any suitable range of greyscale values.
Although the use of infrared radiation having a predetermined wavelength of 1550nm is preferred, infrared radiation having a predetermined wavelength of between 1400nm and 1600nm could be emitted by the laser 23 and captured by the infrared camera 16 to provide the captured infrared image. This is because most modem vehicle windscreens employ materials, such as coatings or layers, which prevent the transmission of infrared radiation through the windscreen into the vehicle interior 15. These materials are generally provided to prevent the vehicle interior from heating up in sunny conditions, but may also prevent the operation of imaging apparatus which relies on the use of infrared radiation to capture infrared images.
The inventors have, however, appreciated that the transmission of infrared radiation having a wavelength within the range of approximately 1400nm to approximately 1600nm is not prevented by such materials and, hence, an accurate infrared image of the vehicle interior 15 can be captured using the imaging apparatus 10 according to the invention which utilises infrared radiation within this range.
It is possible that objects inside the vehicle 14, other than human skin, may contain water and therefore absorb infrared radiation at the predetermined wavelength. For example, natural vegetation such as a cabbage may contain water and, therefore, appear as a dark area in the captured infrared image.
Since such objects which appear as dark areas in the captured infrared image may be of a size similar to a human head, it is undesirable to rely solely on the captured infrared image to determine the number of occupants in the vehicle 14 since it would be possible to easily fool the apparatus 10 and, for example, avoid the imposition of penalties in situations such as those where the apparatus 10 is used to determine the number of occupants in vehicles using high occupancy vehicle lanes.
In view of the above, in order to improve the accuracy of occupant detection using the imaging apparatus 10, as indicated above a shorter wavelength infrared image of the common view 19 of the vehicle interior 15 is also captured using the video camera 17. In the preferred embodiment of the invention, an infrared image in which the wavelength of the infrared radiation is in the range of the electromagnetic spectrum between approximately 800 nm and 900 nm, is captured using the video camera 17.
The inventors have appreciated that restriction of the wavelength of the infrared radiation to this small range in the captured video image is advantageous because all skin reflects radiation in this range, irrespective of
skin type or colour. The inventors have also appreciated that a standard video camera is sensitive to this specific portion of the infrared spectrum, and thus can be used as the means of image capture, rather than a dedicated infrared camera.
The imaging apparatus 10 also includes a second infrared illumination means, such as a laser 24. The laser 24 provides infrared illumination having a predetermined wavelength in the range 800 nm to 900 nm to the vehicle interior 15. At this wavelength range there is risk of retinal damage especially to anyone not protected by the infrared absorbing layers of a modern car windscreen. The beam spreading device (not shown) for laser 24 thus contains a diffusing element to ensure the eye safety of anyone not suitably protected. In an embodiment therefore, separate optical arrangements, rather than a common optical arrangement (such as 18 for the image capture) are used to deliver the illumination from the two illumination means to the interior of the vehicle 15.
In order to ensure that the video camera 17 receives infrared radiation in the predetermined wavelength band only, the imaging apparatus 10 includes a filter device 25, for example a notch filter or band pass filter which matches the wavelength provided by laser 24 (e.g. 808 nm), which filters the infrared radiation provided by the optical arrangement 18 to the video camera 17.
The use of radiation in the shorter region of the infrared spectrum provides a captured video greyscale image comprising dark and light areas, having different greyscale values, defined by individual pixels in the video image. Areas of the captured video image in which exposed skin is present may appear darker or lighter than areas in which no exposed skin is present due to the fact that the surrounding artificial materials and objects may reflect the incident infrared radiation to a greater or lesser extent than exposed skin.
Due to the fact that exposed skin reflects incident shorter wavelength infrared radiation, it will always appear lighter in the captured video image than in the captured infrared image. In contrast green organic material such as a cabbage absorbs in both wavelength ranges and therefore appears dark in both images.
However, areas of the captured video image in which no exposed skin is present will, in most cases, have a similar appearance to corresponding areas of the captured infrared image in which no exposed skin is present. This is because objects inside the vehicle 14, for example hair and artificial materials such as clothing and upholstery, will generally reflect incident infrared radiation having a wavelength in the range of 800 to 900nm, and incident infrared radiation having a wavelength of approximately 1550nm, in a similar way.
Furthermore, in order to maximise the similarity of the appearance of corresponding areas of the captured infrared and video images in which no exposed skin is present, the processor 21 is operable to continuously monitor the greyscale values of pixels defining one or more selected predetermined areas in each of the infrared and video images in which no exposed skin is present, and to adjust or calibrate the greyscale values of these predetermined areas so that they are substantially the same in both the infrared and video images.
Once the infrared and video images of the common view 19 have been captured, the processor 21 is operable to analyse the captured images by comparing the images to detect differences between the images. Based on the detected differences, the processor is then able to determine the number of occupants in the vehicle 14.
The processor 21 analyses the captured infrared and video images by mathematically processing the greyscale values of corresponding individual
pixels in the captured infrared and video images. In one embodiment of the invention, the processor 21 analyses the captured images by subtracting the greyscale values of individual pixels in the captured infrared image from the greyscale values of corresponding individual pixels in the captured video image to provide a processed composite image of the common view 19 of the vehicle interior 15.
Referring to Fig. 3, as explained above, the greyscale values of individual pixels in the captured infrared image 33 of the common view 19 will be low in areas where exposed skin, such as the exposed skin of a human face 31 , is present so that these areas will appear dark, and generally black. In the example of Fig. 3, the pixels defining the area of the captured infrared image 33 of the common view 19 in which the human face 31 is present have greyscale values of 20. This is due to the absorption of substantially all of the infrared radiation at the predetermined wavelength of 1550nm by the exposed skin.
In the captured video image 32 of the common view 19, the greyscale values of the corresponding individual pixels defining the area in which the exposed skin of the human face 31 is present will always be greater due to the fact that the exposed skin reflects incident radiation in this portion of the infrared range, again as explained above. In the example of Fig. 3, the pixels defining the area of the captured video image 32 of the common view 19 in which the human face 31 is present have greyscale values of 100.
The average greyscale values of the corresponding pixels defining corresponding areas of the captured infrared and video images 33, 32 in which no exposed skin is present are similar, as explained in detail above. In the example shown, the average greyscale value of the pixels in the areas of the captured infrared 33 image in which no exposed skin is present is 78, whilst in the captured video image 32 the average greyscale value of the corresponding pixels is 80.
Subtraction of the grayscale values of individual pixels in the captured infrared image 33 from the greyscale values of the corresponding individual pixels in the captured video image results in a processed composite image 34 in which the individual pixels defining the area of the common view 19 in which the human face 31 is present have greyscale values of 80, and in which the individual pixels defining the area of the common view 19 in which no exposed skin is present have an average greyscale value of 2. The area of the processed composite image 34 in which the human face 31 is present will therefore appear lighter than the other areas of the processed composite image 34 which will appear darker.
The processed composite image 34 is then inverted, for example the processor 21 may be operable to invert the image, so that the lighter area which corresponds to the human face 31 becomes darker, whilst the other areas of the processed composite image 34 become lighter. Areas of the inverted processed composite image 34 in which exposed skin is present will therefore appear much darker than the surrounding areas in which no exposed skin is present, thus enabling vehicle occupants to be readily distinguished from the background and readily identified in the processed composite image 34.
Under certain conditions, especially if there is additional ambient illumination, it is possible for the distinction between the areas of exposed skin 31 in the inverted composite image and the background to be much smaller than described above. In a second (preferred) embodiment of the invention, the processor 21 analyses the captured images by performing an AND operation which matches dark areas (i.e. low greyscale values) of the captured infrared image 33 corresponding to possible human skin 31 absorption with pale areas (i.e. high greyscale values) in the captured video image 32 corresponding to possible human skin 31 reflection. In this way both images have to indicate the presence of human skin 31 in the common view 19 of the
vehicle interior 15 in order for the imaging apparatus 10 to report such presence.
As any single occupant in the vehicle 14 will have a number of body parts with exposed skin, for example face, hands, etc, the composite processed image is likely to comprise a plurality of darker areas corresponding to exposed skin.
In order to determine the number of occupants in the vehicle, the processor 21 is operable to analyse the processed images to determine which of the areas in the processed image corresponding to human skin are in fact human faces. In particular, the processor 21 is operable to filter the 'skin' areas in the processed image, for example based on the size of the areas, or the aspect ratios of the areas, both of which can be correlated with known sizes and aspect ratios for human faces, to identify the 'skin' areas which correspond to human faces. The processor 21 is finally operable to count the number of
'skin' areas identified as corresponding to human faces to determine the number of occupants in the vehicle 14.
Depending upon the situation in which the imaging apparatus 10 is employed, the output from the apparatus 10 may be used to trigger further apparatus. For example, if the processor 21 determines that there is only one occupant in a vehicle travelling in a high occupancy vehicle lane where the minimum number of permitted occupants is two, the imaging apparatus 10 may trigger further apparatus such as a standard enforcement camera to capture an image of the vehicle from which the driver and vehicle registration plates can be easily identified for subsequent enforcement purposes.
A particular advantage of the imaging apparatus 10 according to the present invention is that by employing the optical arrangement 18, a common view 19 of the vehicle interior 15 can be provided to both the infrared and video cameras 16, 17. This ensures that the individual pixels in each of the captured
infrared and video images 33,32 provide a view of an identical part of the vehicle interior 15. The pixels in the separate infrared and video images are thus fully aligned such that there is no parallax between them, which could arise if cameras 16, 17 with individual lenses were employed to capture 5 images of the vehicle interior 15.
In order to further ensure that the infrared and video cameras 16, 17 both capture an identical image of the common view 19 of the vehicle interior 15, the sensors (also known as camera chips) within the two cameras are
I O preferably matched both pixel for pixel, and dimensionally. This ensures that the fields of view of the infrared and video cameras 16, 17 are matched, regardless of the distance between each of the cameras 16, 17 and the focal plane of the common view 19. This also ensures that the magnification of the common view 19, provided by the infrared and video cameras 16, 17, is
15 exactly the same. However, it is not always possible to achieve this, and adjustments then have to be made in software to allow for the differences.
In a preferred embodiment of the imaging apparatus 10, the optical arrangement 18 includes a lens arrangement having a single lens 26 which is
20 focussed on the vehicle interior 15. It may not be possible using the imaging apparatus 10 to obtain a complete view of the vehicle interior 15 due to the fact that some of the rear interior of the vehicle 14 may be obscured by the front seats and the roof. However, by directing the lens 26 towards the front and one side of the vehicle 14, a common view 19 of the front seats and part of the rear
25 seats can be provided to the infrared and video cameras 16, 17 which is considered to be sufficient for most situations. If a complete view of the vehicle interior 15 is required, it will generally be necessary to employ two suitably positioned imaging systems 10.
30 In order to enable the lens 26 to transmit the common view 19 of the vehicle interior 15 to both the infrared and video cameras 16, 17, the optical arrangement 18 also includes a beam splitter device 27. The beam splitter
device 27, which may comprise coated glass, is adapted to allow the transmission therethrough of infrared radiation at the predetermined wavelength to direct the infrared radiation from the lens 26 into the infrared camera 16 to enable it to capture an infrared image of the common view 19. Infrared radiation received from the lens 26 in the wavelength range 800 to 900nm is reflected by the beam splitter device 27 into the video camera 17 to enable the video camera 17 to capture a video image of the common view 19.
Although embodiments of the invention have been described in the preceding paragraphs with reference to various examples, it should be appreciated that various modifications to the examples given may be made without departing from the scope of the present invention, as claimed.
For example, infrared image capture means other than infrared and video cameras 16, 17 may be used. An alternative optical arrangement 18 to a lens 26 or beam splitter device 27 may be used to convey the infrared images of the common view 19 to the infrared and video cameras 16, 17. The beam splitter device 27 may be arranged to direct infrared radiation received from the lens 28 into the infrared camera 16 by reflecting it into the infrared camera 16, and may be arranged to direct infrared radiation received from the lens 28 into the video camera 17 by allowing the transmission of the infrared radiation through the beam splitter device 27 into the video camera 17. The imaging apparatus 10 may utilise a range of greyscale values between 0 and 1023, although as mentioned above, any suitable range of greyscale values may be employed. The processor 21 may be operable to analyse or compare the captured infrared and video images other than by subtracting the greyscale values of corresponding individual pixels in the respective images, or by performing an AND function. For example, the processor 21 may be operable to mathematically process the greyscale values of corresponding individual pixels in the captured infrared and video images. This mathematical processing could, for example, be by addition or multiplication of the greyscale
values of corresponding individual pixels in the captured infrared and video images.
Whilst endeavouring in the foregoing specification to draw attention to those features of the invention believed to be of particular importance, it should be understood that the Applicant claims protection in respect of any patentable feature or combination of features hereinbefore referred to and/or shown in the drawings, whether or not particular emphasis has been placed thereon.
Claims
1. Imaging apparatus for determining the number of occupants in a vehicle, comprising: means for illuminating an interior of the vehicle within a first range of infrared wavelengths; a first capture means for capturing a first infrared image of the interior for the first range ϋf wavelengths; means for illuminating the interior of the vehicle within a second range of infrared wavelengths; a second capture means for capturing a second infrared image of the interior for the second, range of wavelengths; and a processor for analysing the first and second captured infrared images ϋf the interior of the vehicle and operable to determine the number of occupants in the vehicle.
2. Imaging apparatus according to Claim 1, wherein the second range of infrared wavelengths is between approximately δOOnm and approximately 900nm.
3. Imaging apparatus according to Claim 2, wherein the means for illuminating the interior within the second range of wavelengths is adapted to illuminate the interior with radiation at approximately 808nm.
4. Imaging apparatus according to any preceding claim, wherein the second capture means is a conventional video capture means.
5. Imaging apparatus according to Claim 4, wherein the second capture means is a video camera.
6. Imaging apparatus according to any preceding claim, wherein the first range of infrared wavelengths is between approximately 1400nm and approximately 1600nm.
7. Imaging apparatus according to Claim 6, wherein the means for illuminating the interior within the first range of wavelengths is adapted to illuminate the interior with radiation at approximately 1550nm.
8. Imaging apparatus according to any preceding claim, wherein the first and second capture means include narrow band filter devices matching the respective wavelengths of the first and second illuminating devices.
9. Imaging apparatus according to any preceding claim, wherein the processor is operable to perform an AND function on the first and second captured infrared images.
10. A method for determining the number of occupants in a vehicle, the method comprising: illuminating an interior of the vehicle within a first range of infrared wavelengths; capturing a first infrared image of the interior for the first range of wavelengths; illuminating the interior of the vehicle within a second range of infrared wavelengths; capturing a second infrared image of the interior for the second range of wavelengths; and analysing the first and second captured infrared images of the interior of the vehicle to determine the number of occupants in the vehicle.
11. The method of Claim 10, wherein the second range of infrared wavelengths is between approximately 800nm and approximately ΘOOnm.
12. The method of Claim 11, comprising illuminating the interior within the second range of wavelengths with radiation at approximately 80δnm.
13. The method of any of the Claims 10 to 12, comprising capturing the second infrared image with a conventional video capture means.
14. Imaging apparatus for determining the number of occupants in a vehicle, comprising a means of capturing a first infrared image together with a means of illuminating the interior of the vehicle at the first infrared wavelength, a means of capturing a second infrared image together with a means of illuminating the interior of the vehicle at the second infrared wavelength, an optical arrangement adapted to provide a common view of the interior of the vehicle to both means of infrared image capture to enable the capture of two infrared images at different wavelengths of the common view of the interior of the vehicle, and a processor to analyse the captured infrared images of the common view of the interior of the vehicle and able to determine the number of occupants in the vehicle.
15. A method for determining the number of occupants in a vehicle, the method comprising: providing a common view of an interior of a vehicle to both means of capturing infrared images; capturing infrared images of the common view of the interior of the vehicle using two means of capturing infrared images respectively; analysing the captured infrared images of the common view of the interior of the vehicle to detect the presence of persons in the captured ima'ges; and determining the number of occupants in the vehicle based on the image analysis.
16. The method of Claim 15, further comprising illuminating the interior of the vehicle at the first infrared wavelength, and illuminating the interior of the vehicle at the second infrared wavelength.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| GBGB0703031.5A GB0703031D0 (en) | 2007-02-15 | 2007-02-15 | Method and apparatus for counting vehicle occupants |
| GB0703031.5 | 2007-02-16 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2008099146A1 true WO2008099146A1 (en) | 2008-08-21 |
Family
ID=37908748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/GB2008/000460 WO2008099146A1 (en) | 2007-02-15 | 2008-02-12 | Method and apparatus for counting vehicle occupants |
Country Status (2)
| Country | Link |
|---|---|
| GB (1) | GB0703031D0 (en) |
| WO (1) | WO2008099146A1 (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011128668A2 (en) | 2010-04-13 | 2011-10-20 | Vehicle Occupancy Limited | Improvements in or relating to vehicle occupant detectors |
| CN103164708A (en) * | 2011-12-13 | 2013-06-19 | 施乐公司 | Determining a pixel classification threshold for vehicle occupancy detection |
| US9477881B2 (en) | 2012-10-19 | 2016-10-25 | Nec Corporation | Passenger counting system, passenger counting method and passenger counting program |
| WO2017158648A1 (en) * | 2016-03-17 | 2017-09-21 | 日本電気株式会社 | Passenger counting device, system, method and program |
| FR3111449A1 (en) | 2020-06-15 | 2021-12-17 | Fareco | Systems and methods for determining the number of people in a vehicle |
| WO2023028714A1 (en) * | 2021-09-02 | 2023-03-09 | Invision Ai, Inc. | Vehicle occupancy detection system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6370260B1 (en) * | 1999-09-03 | 2002-04-09 | Honeywell International Inc. | Near-IR human detector |
| US20050111703A1 (en) * | 2003-05-14 | 2005-05-26 | Peter-Michael Merbach | Method and apparatus for recognition of biometric data following recording from at least two directions |
| WO2005096216A1 (en) * | 2004-03-29 | 2005-10-13 | Siemens Corporate Research, Inc. | Face detection and recognition using pose-specific detectors and pose-specific recognizers |
| WO2005119571A1 (en) * | 2004-06-04 | 2005-12-15 | Laser Optical Engineering Limited | Imaging apparatus and method for vehicle occupant determination |
-
2007
- 2007-02-15 GB GBGB0703031.5A patent/GB0703031D0/en not_active Ceased
-
2008
- 2008-02-12 WO PCT/GB2008/000460 patent/WO2008099146A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6370260B1 (en) * | 1999-09-03 | 2002-04-09 | Honeywell International Inc. | Near-IR human detector |
| US20050111703A1 (en) * | 2003-05-14 | 2005-05-26 | Peter-Michael Merbach | Method and apparatus for recognition of biometric data following recording from at least two directions |
| WO2005096216A1 (en) * | 2004-03-29 | 2005-10-13 | Siemens Corporate Research, Inc. | Face detection and recognition using pose-specific detectors and pose-specific recognizers |
| WO2005119571A1 (en) * | 2004-06-04 | 2005-12-15 | Laser Optical Engineering Limited | Imaging apparatus and method for vehicle occupant determination |
Non-Patent Citations (2)
| Title |
|---|
| IOANNIS PAVLIDIS ET AL: "A Vehicle Occupant Counting System Based on Near-Infrared Phenomenology and Fuzzy Neural Classification", IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, IEEE, NEW YORK, NY, US, vol. 1, no. 2, June 2000 (2000-06-01), XP011028372, ISSN: 1524-9050 * |
| KONG S G ET AL: "Recent advances in visual and infrared face recognition-a review", COMPUTER VISION AND IMAGE UNDERSTANDING, ACADEMIC PRESS, SAN DIEGO, CA, US, vol. 97, no. 1, January 2005 (2005-01-01), pages 103 - 135, XP004642050, ISSN: 1077-3142 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011128668A2 (en) | 2010-04-13 | 2011-10-20 | Vehicle Occupancy Limited | Improvements in or relating to vehicle occupant detectors |
| CN103164708A (en) * | 2011-12-13 | 2013-06-19 | 施乐公司 | Determining a pixel classification threshold for vehicle occupancy detection |
| US9846802B2 (en) | 2012-10-19 | 2017-12-19 | Nec Corporation | Passenger counting system, passenger counting method and passenger counting program |
| US9477881B2 (en) | 2012-10-19 | 2016-10-25 | Nec Corporation | Passenger counting system, passenger counting method and passenger counting program |
| US20190205677A1 (en) * | 2016-03-17 | 2019-07-04 | Nec Corporation | Passenger counting device, system, method and program |
| US10318830B2 (en) | 2016-03-17 | 2019-06-11 | Nec Corporation | Passenger counting device, system, method and program |
| WO2017158648A1 (en) * | 2016-03-17 | 2017-09-21 | 日本電気株式会社 | Passenger counting device, system, method and program |
| US10789494B2 (en) | 2016-03-17 | 2020-09-29 | Nec Corporation | Passenger counting device, system, method and program |
| US10824887B2 (en) | 2016-03-17 | 2020-11-03 | Nec Corporation | Passenger counting device, system, method and program |
| US10922565B2 (en) | 2016-03-17 | 2021-02-16 | Nec Corporation | Passenger counting device, system, method and program |
| FR3111449A1 (en) | 2020-06-15 | 2021-12-17 | Fareco | Systems and methods for determining the number of people in a vehicle |
| EP3926529A1 (en) | 2020-06-15 | 2021-12-22 | Fareco | Systems and methods for determining a number of persons in a vehicle |
| WO2023028714A1 (en) * | 2021-09-02 | 2023-03-09 | Invision Ai, Inc. | Vehicle occupancy detection system |
| US11804058B2 (en) | 2021-09-02 | 2023-10-31 | Invision Ai, Inc. | Road side vehicle occupancy detection system |
Also Published As
| Publication number | Publication date |
|---|---|
| GB0703031D0 (en) | 2007-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4442472B2 (en) | Device part identification device | |
| US20190243092A1 (en) | Imaging apparatus, imaging system that includes imaging apparatus, electron mirror system that includes imaging apparatus, and ranging apparatus that includes imaging apparatus | |
| US7280678B2 (en) | Apparatus and method for detecting pupils | |
| CN100376922C (en) | Personnel identifier and fingerprint image camera method | |
| US6603137B2 (en) | Differential imaging rain sensor | |
| US8254768B2 (en) | System and method for illuminating and imaging the iris of a person | |
| JP6333238B2 (en) | Raindrop detection on glass surface using camera and lighting | |
| KR101433715B1 (en) | System and method for robust fingerprint acquisition | |
| KR20010032372A (en) | Method and apparatus for illuminating and imaging eyes through eyeglasses using multiple sources of illumination | |
| WO2008099146A1 (en) | Method and apparatus for counting vehicle occupants | |
| KR20120116882A (en) | Determining a number of objects in an ir image | |
| US20150062319A1 (en) | Authentication apparatus, prism member for authentication and authentication method | |
| JP2006351011A (en) | Color imaging system for locating retroreflector | |
| JP2000028315A (en) | Object detector | |
| US7091867B2 (en) | Wavelength selectivity enabling subject monitoring outside the subject's field of view | |
| US9619690B2 (en) | Authentication apparatus, prism member for authentication and authentication method | |
| CN108710844B (en) | Authentication method and device for detecting face | |
| US20150054932A1 (en) | Authentication apparatus, prism member for authentication and authentication method | |
| WO2005119571A1 (en) | Imaging apparatus and method for vehicle occupant determination | |
| US8831416B2 (en) | System and method for illuminating and identifying a person | |
| US20130093891A1 (en) | Vehicle occupant detectors | |
| CN112782090B (en) | Drunk driving automatic monitoring system and detection method | |
| Tyrer et al. | An optical method for automated roadside detection and counting of vehicle occupants | |
| CN112622919B (en) | Drunk driving automatic monitoring system and detection method | |
| GB2506688A (en) | Detection of a target in a scene using hyperspectral imaging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 08709357 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 08709357 Country of ref document: EP Kind code of ref document: A1 |