The present invention relates to thin-film cameras and other optical devices, and in particular to capturing an image from any point within the periphery of a slim display.
Video displays are typically opaque in the sense that a camera placed behind the display cannot capture an image of something in front of the display. Cameras are therefore often placed at the edge of a display with the unwanted result that speakers in a video conference appear not to be looking at one another. It is thought that users would like a display to behave more like a window in the sense that an image could be captured as if from a camera at any point behind the screen.
It is well-known that the compound eye of a house fly covers a large area but is thin. Following this lead, arrays of tiny cameras have been placed in front of screens with the aim of capturing an image from any point on the screen. The resolution of any camera is, however, limited by aperture diffraction: the number of rows or columns in the image can be no greater than the diameter of the camera lens divided by the wavelength of light. The focal length of a camera will typically be at least as big as its diameter so camera arrays are rarely thinner than 1 mm while having the further disadvantage that they are opaque. Note that the information passed by each element of a compound eye or camera can be much less than for a conventional camera: it is the lens of a compound eye, not information theory, that keeps it thick.
WO 02/45413 by the present inventor disclosed a wedge-shaped wave-guide and a turning film placed adjacent to a screen that transferred light from objects in front of the screen to a camera at the screen edge (see
Note, however, that because rays are deflected so as to travel parallel to the guide surface, it is possible to bring to a focus rays from an aperture that is thin but wide, i.e. a slit aperture. Even a very thin system might in principle therefore produce images with a resolution that is high in one dimension albeit low in the orthogonal dimension.
This invention is defined in the claims. Exemplary important features of embodiments are:
1. An array of cameras that are each two-dimensional in the sense of having a one-dimensional aperture (slit) through which rays are brought to focus by a two-dimensional or planar lens onto a one-dimensional array of photosensors.
2. The slits of the cameras are arranged at a variety of angles relative to one another.
3. A computed tomography algorithm is used to calculate the two-dimensional image from the several one-dimensional images formed by the cameras.
For a better understanding of the invention, embodiments will now be described with reference to the attached drawings, in which:
Referring to
At the lens aperture, which is a long (notional) slit 14, there is a prism 12, reflecting strip or similar that deflects light from air into the plane of the lens 16, which being thin acts as a waveguide. Rays are then refracted by the guide 16 to converge at points on a linear photodetector array 20 parallel to the aperture on the opposite side of the lens waveguide.
In
The array of cameras 10a, 10b, 10c will be laid like a film on the surface of a flat panel display 50 and, because the sensor array is only a small fraction of the area of each camera, the area of the camera will be mostly transparent to light from the display. An opaque layer (not shown) may be placed between the sensors and the display in order that light from the display does not affect the sensors.
The image-processing task is like that in X-ray computed tomography where each slice of the object is two-dimensional and each detector captures X-rays that have travelled through one column of the slice, summing the various local intensities throughout the column.
Algorithms for computed tomography are described in S. W. Smith, ‘The Scientist and Engineer's Guide to Digital Signal Processing’, California Technical Publishing, pages 444-449, 1997. A good approach is to take the one-dimensional Fourier transform of the intensities from each camera and plot each Fourier transform in the Fourier plane at an angle to the horizontal that equals the angle to the display horizontal of the associated camera. This is illustrated in
In a typical use, the object is to be placed in front of the display; normally the object will be the user, who is looking at a particular part of the screen, but generally around the middle. For a small screen such as that on a mobile phone, or even a large screen, this is already an improvement, as compared to a camera located off the edge of the screen.
For further improvement, the entire screen, or most of it, or at least 70%, 80% or 90%, may be covered by the cameras and the image processor may use a closely spaced group of two-dimensional cameras from whichever point of view on the display is desired, which may change, for example, during a conference call with several participants.
It is desirable that the array of slits handle light as efficiently as a conventional camera; a 33 mm by 33 mm array of slits measuring 1 mm by 1 micron (i.e. about 1000 slits) will have the same surface area as a conventional camera with an aperture of 1 mm by 1 mm. However, aperture diffraction by the slit may cause much of the transient light to miss the line of photodetectors. The lens plane should therefore comprise a wave-guide with sufficient numerical aperture to collect all light passing through the slit. If the guide is multi-mode, rays at higher order modes may be focused by the lens over a different distance than rays in the fundamental mode. Preferably, therefore, the guide will be monomode; for example, it may be a step-index guide.
The disclosure thus concerns a planar optical element, in particular a camera, comprising a diverter 14 for diverting light from an object into an imaging plane; a planar lens waveguide 16 in the imaging plane, receiving the diverted light and focusing it onto a line; and a sensor line 20 located on the focus line, for forming a one-dimensional image of the object. Many such elements can be applied to a planar substrate, such as a display screen, at different angles to a given direction on the substrate, and the one-dimensional inputs Fourier-analysed to reconstruct the desired two-dimensional image.
The elements (apart from the diverter) can be transparent to light, so that the substrate can be a display screen; this eliminates the need to locate a camera to the side of the screen, and also means that for videoconferencing the user looking at the screen will be looking into the camera. For larger screens the elements can cover all or most of the screen, and a subset of the elements, covering a relatively small area of, say, 30×30 mm, chosen at any given time to constitute the camera.
The system can also be run backwards as a projector, with light-emitting elements instead of sensors.
Number | Date | Country | Kind |
---|---|---|---|
21306664.0 | Nov 2021 | EP | regional |