Image analysis – Applications – 3-d or stereo imaging analysis
Reexamination Certificate
2000-03-31
2004-09-21
Mariam, Daniel (Department: 2621)
Image analysis
Applications
3-d or stereo imaging analysis
C356S012000, C348S042000, C348S049000, C348S050000
Reexamination Certificate
active
06795575
ABSTRACT:
This invention relates to line-scan imaging in 3-D, and in particular to the use of an imaging detector such as a single CCTV camera.
Line-scan cameras have been in existence for many years. They consist of a single line of photo sites, usually made of silicon, in front of which is a standard lens controlling focus, aperture etc. This is effectively a one dimensional sensor which can form the image of a scene if relative lateral movement takes place between the object of interest and the sensor. The line-by-line sequence of imaging information is built up into an area image by using a frame store. Typical applications are in checking components being moved by conveyor in a manufacturing environment. These components are usually flat, such as printed circuit boards, where 2-D pattern matching algorithms are used for inspection purposes.
A 3-D (stereoscopic) line-scan system could be built using two line-scan cameras configured to provide left and right perspective information. With such a system it would be possible to extract 3-D coordinate data from the image of an object and thus make 3-D inspection possible. For such a system to work there would need to be precise geometrical alignment of both line-scan cameras and also well matched lenses would be essential. These requirements could cause severe problems in setting up such a system since an image could only be formed after movement has taken place and so real-time adjustments of such simple things as focus are not possible. Since precise matching between the parameters of both cameras is essential if acceptable 3-D images are to be produced, then the set-up procedure can be both time consuming and imprecise.
It is, therefore, an object of the invention in one aspect to avoid these problems.
Therefore, according to the invention apparatus is provided for producing a stereoscopic line-scan image of an object in which a detector camera, such as a television camera, is used to view the object, the detector camera having an output signal generated by scanning the received image in a series of parallel lines across the detector, means are provided to move relative to one another the surface of the object past the area viewed by the detector in a direction which is at right angles to the parallel lines across of the detector, and means are provided to select the outputs from two spaced lines in each picture frame produced by scanning the detector and store those lines in a frame store to build up individual pictures from each of the two lines, so that one individual picture constitutes a left hand picture of a stereoscopic pair and the other individual picture constitutes a right hand image of a stereoscopic pair.
Also according to the invention there is provided a method of producing a stereoscopic pair of images, in which the surface of an object is moved past a detector camera which produces an output signal generated by scanning the received image in a series of parallel lines across the detector, the direction of movement of the object being at right angles to the scanning line direction of the detector, and the outputs from two spaced lines in each picture frame produced by the detector are stored and individual pictures are built up from each of the two lines, so that one individual picture constitutes a left hand picture of the stereoscopic pair and the other individual picture constitutes a right hand image of the stereoscopic pair.
By operating in this way all of the problems of the set up of two separate cameras or detectors are avoided. Thus one only selects the image information from two of the multiplicity of video lines available each frame. If the detector is a camera it can be used as a standard television camera to view the object during set up and its direction, focus, brightness and contrast can be selected to give a good, high quality, image. Then, one can switch to the arrangement according to the invention where the video information from single lines is selected and, as the surface of the object is moved past the television camera, successive signals from those individual lines are selected, stored and then put together to form the two separate pairs of stereoscopic images. All of this can be achieved by means of a suitable program and stored in a frame store and thereafter the image can be looked at and manipulated as required.
By way of example, the left and right hand images which result can be viewed as a stereoscopic pair and there are very well known systems available such as the red/green anaglyph, field sequential, and lenticular screen systems, which enable one to present or view left and right hand images. Therefore any of these systems can be used.
In a more sophisticated operation as will be described below, however, one can subtract the video signal in the left hand image from that in the right hand image, or vice versa. This will have the effect of providing an output signal which is unique. Therefore, one can readily use and store that signal in a data base and compare it with signals obtained from other objects to see if there is correspondence. For example such a technique can be extremely useful in forensic analysis in the comparison of cylindrical objects like spent cartridge cases or in the examination of the bullets after firing as will be described in more detail below.
In order to enhance or reduce the depth of the resulting stereoscopic image when viewed one can, of course, chose lines having varying positions in the overall video picture frame. By choosing lines which are spaced apart as far as possible across the video frame, i.e. a line from the very beginning of each complete scan and a line at the end of such a scan, one can create the greatest impression of depth. By choosing lines which are closer together a reduced impression of depth can be given. This can be achieved by programme means which processes the video signal so that an operator can chose a particular impression of depth which he find suits his purposes the best. The system of the invention is, therefore, very flexible and can present information to the unique requirements of any particular operator yet in a very simple fashion.
Where the detector camera is a television camera, it is being used according to this aspect of the invention, for two distinct functions and such a use is quite unique. Thus the video signal from the television camera can be used as a standard video signal and displayed in standard fashion on a monitor. Alternatively the signal can be processed according to the invention to select individual lines in a scan and the results used to build up a pair of stereoscopic images which can then be viewed or presented stereoscopically.
In another embodiment of the invention X-ray images can be obtained using a point source X-ray and as the imaging chip a full-field digital X-ray plate detector. Such detectors can give an output analogous to the raster scan output from a conventional television camera, the only difference between this arrangement and the use of visible light being that a shadow picture is provided by the X-ray source rather than a light image. The output, however, is a digital signal and individual lines of the signal can be selected to build up the individual picture.
The direction of scanning of the detector must be at right angles to the direction at which the surface of the object moves past it. In normal circumstances with a conventional television camera, this will tend to mean that the scanning lines need to be vertically oriented so that it can look at the surface of an object moving past it horizontally. Traditionally a television camera has horizontal scanning lines and so in one very simple embodiment according to the invention one simply turns the camera through 90° so that it's scanning lines are then substantially upright. Alternatively, however, it is possible to produce by program means upright or vertical lines from the video signal, in which case it would then not be necessary to reorient the camera.
In a particularly favourable embodiment according to the inven
Evans Paul
Robinson Max
Fredrikson & Byron , P.A.
Mariam Daniel
Nottingham Trent University
Patel Shefali
LandOfFree
Line-scan imaging in 3-D does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Line-scan imaging in 3-D, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Line-scan imaging in 3-D will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3208448