Classifications

• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T7/00—Image analysis
  • G06T7/50—Depth or shape recovery
  • G06T7/55—Depth or shape recovery from multiple images
  • G06T7/593—Depth or shape recovery from multiple images from stereo images
• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T7/00—Image analysis
  • G06T7/70—Determining position or orientation of objects or cameras
  • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods

Definitions

• the invention concerns a method for determining the position of a point, whereby this point is perceived by at least two separate cameras with a number of pixels, where the image of the said point is detected and then the position of that point is calculated on the basis of the position of the pixels so identified.
• each pixel of a camera must be examined in turn to see whether the image of this point falls onto it. Since the cameras used must be of a high resolution, for example 2000x2000 pixels, the search for the image of the said point requires highly complex and expensive equipment, and is also extremely time-consuming.
• the invention is designed to overcome these disadvantages by putting forward a method that enables the position of the point to be determined rapidly and economically and with a high degree of accuracy.
• an address is assigned to each of the pixels of at least one camera from which a group of pixels is selected containing at least some of the pixels where the said image is detected, and in such a way that when the point is moved, the group of pixels is modified so that it continues to contain the pixels where the said image is detected.
• the centre of the image of the said point is calculated by constituting the said group of a fixed number of pixels and then modifying it so that its centre lies at the centre of the image of the said point as calculated.
• an imaginary window is selected containing the said group and the path of the image of the said point is followed with this window.
• the invention also includes a device for establishing the position of a point, using at least two separate cameras, and means for determining the position of the said point on the basis of the image formed by each of the cameras.
• a feature of this device is that at least one of the said cameras is capable of enabling each of the pixels to be identified separately.
• means are provided to select a group of pixels from among the pixels of a camera, in such a way as to ensure that the group contains the pixels where the image of the said point is detected, and whereby means are provided to modify the said group when the said point is moved so that it continues to contain the pixels where the said image is detected.
• At least one of the said cameras is a so-called "addressable" camera.
• Figure 1 is a schematic representation of a device developed by the invention.
• Figure 2 is a schematic representation of the screen of a camera.
• the method and the device covered by the invention constitute a measuring system, and in particular a so-called "optical system for measuring coordinates", in which the co-ordinates of a point are measured with respect to a reference.
• the point to be measured should consist of a point of light or an element that emits an optical signal such, for example, as a lamp. In this way it is possible, for instance, to determine the distance between two points or to record the shape of an object. This data can be stored in the memory of a computer for subsequent processing.
• Figure 1 shows two cameras (cameras 1 and 2) that are aimed at a point identified by a point of light 3. These cameras are positioned at an angle to each other and their position relative to each other is known. The resolution of these cameras is determined, among other things by the number of pixels on which an image of an object can be detected. It is thus possible, when the position of the image of light-point 3 on cameras 1 and 2 is known precisely, to calculate using conventional trigonometry the correct position or the correct co-ordinates of the said light-point 3 in relation to these cameras or in relation to a reference point fixed in relation to cameras 1 and 2.
• Window 4 is chosen so that it contains the pixels of camera 1 or 2 on which an image 5 of the light-point 3 is formed. Ideally the centre of window 4 is made to lie at the centre of the pixels holding image 5 of the light-point 3.
• the processing time of such a measuring system is significantly reduced, since only a limited number out of the total number of pixels has to be used to determine the position of the said light-point 3. In other words, the data to be processed is drastically reduced.
• window 4 shifts in such a way as to ensure that image 5 continues to stay within it.
• the group of pixels that constitute window 4 is modified in such way that the pixels on which the image 5 is formed, continue to remain a part of that group.
• An imaginary window 4 is thus generated that follows the path of the movement of image 5 in line with the movement of light-point 3. This is schematically illustrated in Figure 2 where, when light-point 3 moves, image 5 moves in the direction of arrow 6 to an image 5'. In the same way, the position of window 4 shifts to window 4'.
• window 4 In order to shift window 4 or modify the corresponding group of pixels, a calculation will first be made, for instance, of the centre of the pixels on which an image 5 of the said light-point 3 is formed. Window 4 is then moved or the group of pixels modified so that their centre coincides with the centre of the said image 5.
• window 4 is, for instance, enlarged until it once again covers image 5.
• window 4 returns to its original size, doing so in such a way that the image 5 formed remains totally within its confines.
• the method developed by the invention thus combines a high image resolution (of, say, 2,000 x 2,000 pixels) with a high pixel read-out speed that enables extremely accurate measurements to be made.
• Read-out speed means the speed with which data from a specific pixel is available for processing by, say, a computer 8. This high speed is attained because each pixel is assigned its own address.
• the pixels are grouped into specific zones, the same address being assigned to all the pixels within the same zone.
• a zone thus consists, for example, of a column of pixels.
• the method described above is adapted to this approach, processing the said zone as a single pixel.
• the device forming part of the invention and enabling the position of one point - or more specifically the co-ordinates of that point - to be measured with respect to a reference is illustrated schematically in Figure 1.
• Figure 1 This shows two cameras (1 and 2) aimed at a point of light 3, and the means 8 for determining the position of that light-point 3, on the basis of the image captured by each of the cameras 1 and 2.
• the said unit should ideally include a computer running specially adapted software.
• At least one of cameras 1 and 2 contains means enabling each pixel of camera 1 or 2 to be identified separately. It is thus possible, with the aid of computer 8, to receive a signal from a predefined pixel indicating whether an image of light-point 3 is formed on that pixel.
• the device forming part of the invention also enables a zone of pixels to be identified, such, for example, as a column of pixels.
• At least one of the two cameras is an addressable camera.
• each of the two cameras it is preferable for each of the two cameras to be an addressable camera.
• the means to do that consists of a computer 8 with computer equipment.
• Computer 8 also ensures that, when the said point is ed. the said group - or, in other words, window 4 - is modified in such a way that it continues to contain the pixels where image 5 is detected.
• the device developed by the invention is also equipped with a point of light 3.
• This light-point 3 takes the form for example of a lamp mounted on a holder with a pointed end that is held against an object at a spot whose position has yet to be determined.
• Said cameras 1 and 2 and computer 8 are used to determine the positions of one or, preferably, two or more lamps that are fixed to the holder and are thus at a known distance from that pointed end. Moreover, on the basis of the detected position of the said lamps, it is possible to calculate the correct co-ordinates of the spot at which the pointed end of the said holder is located.
• the said light-point 3 takes the form of a light-emitting diode (i.e. a so-called LED).
• the invention is of course not limited to the method described above or to the device illustrated in the annexed figures for determining the position of a point.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Theoretical Computer Science (AREA)
• Length Measuring Devices By Optical Means (AREA)
• Image Analysis (AREA)

Applications Claiming Priority (3)

Publications (1)

Family

ID=3890348

Family Applications (1)

Country Status (4)

Families Citing this family (3)

Family Cites Families (3)

• 1997
  • 1997-02-17 BE BE9700143A patent/BE1010929A3/nl not_active IP Right Cessation
• 1998
  • 1998-02-09 EP EP98901891A patent/EP1101195A1/en not_active Withdrawn
  • 1998-02-09 WO PCT/BE1998/000017 patent/WO1998036381A1/nl active Application Filing
  • 1998-02-09 AU AU58491/98A patent/AU5849198A/en not_active Abandoned

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events