Optical means for making measurements of surface contours

Claims

What is claimed is:

1. A system for creating data to represent a three-dimensional surface shape comprising

means projecting on the surface shape distinctive patterns made up of adjacent pattern portions represented by distinctively differently positioned linear features formed by and between adjacent projected areas having differing optical characteristics, the linear features being positionally located according to a code such that the positions of the linear features projected by at least some of the different patterns are located at different distinct locations such that none of said linear features as projected are at the same location;

image sensor means positioned at a spaced location from the surface shape in position to observe the distinctive projected patterns including the linear features thereof projected on the surface shape, said sensor means including means for producing representations of selected ones of the distinctive linear features formed by the projected pattern portions, and

means for processing at least two image representations of projected patterns projected on the surface shape and sensed by the sensor means, said processing means containing processing data representative of the relative positions in space of the projection means and the image sensor means, said processing means further including means for extracting information from the different image representations to establish data representative of the location in space of different points on the surface shape.

2. The system of claim 1 wherein the detectable linear features are relatively uniformly distributed among the patterns.

3. A system for representing the three-dimensional surface shape of an object, the system including radiant energy projection means for projecting energy containing distinctive patterns of adjacent areas for projection onto the object whose surface shape is to be represented, the patterns defining on the object a plurality of distinctively located profile lines defined by the linear boundary between each pair of adjacent distinctive areas in the projected patterns, the boundaries in each pattern being positioned according to a code such that the locations of the boundaries in at least some of the different patterns are at different distinct locations such that none of the linear boundaries of said projected different patterns will fall on the object at the same location,

image sensor means positioned to observe the patterns projected onto the surface of the object, said sensor means including means for producing representations of selected ones of the projected patterns including the profile lines, and

means for processing the data observed and produced as representations by the sensor means in at least two distinct patterns, said processing means containing processing data representative of the relative positions in space of the projection means and the image sensor means, said processing means further including means for correlating data from the different image representations taking into account the relative positions of the projector and sensor means for producing therefrom data representations of different locations in space that correspond to the locations of profile lines where they occur on the object.

4. The system of claim 3 wherein the linear boundary lines are relatively uniformly distributed among the patterns.

5. A system for use in representing the three-dimensional surface shape of an object, the system including radiant energy projection means, means responsive to energy projected by the projection means and control means for the projection and responsive means, the projection means including a source of energy capable of being controlled to produce projected energy flashes, means to focus radiant energy from the projection means onto an object whose surface shape is to be represented, and an energy transmissivity encoding member mounted in alignment with the energy source and the focusing means, said encoding member having a pattern formed thereon by a plurality of adjacent relatively energy conducting and non energy-conducting areas arranged to extend over a portion of the surface thereof in the region aligned with the energy source and the focusing means, the pattern being segmented into a plurality of discrete segments each having distinctively positioned locations and widths corresponding to the energy conducting and non energy-conducting areas thereof so that regardless of which segment is positioned in alignment with the energy source and its associated focusing means at the time a flash from the energy source is projected, the projected energy will produce a distinctive intensity pattern on the surface of the object defining on the object a plurality of distinctly located profile lines at locations defined by each pair of adjacent energy conducting and non-energy conducting areas on the encoding member.

6. The system of claim 5 wherein the source of energy is a light source and the energy conducting and non energy-conducting areas on the encoding member are substantially light conducting and substantially non light-conducting areas on the encoding member.

7. The system of claim 5 wherein the locations and widths of the light conducting and non light-conducting areas on the encoding member are determined by a code having distinct code portions that determine the locations and widths of each of the light conducting and non light-conducting areas in each segment of the pattern.

8. The system of claim 5 wherein the code is a concatenated binary code having a distinctive sub-portion associated with each pattern segment, each sub-portion having a plurality of positions representing respectively each of the light conducting and non light-conducting areas of each of the corresponding pattern segments.

9. The system of claim 5 including a light tight enclosure having positions therein for mounting an object to be represented, said light enclosure also having spaced positions therein for mounting a plurality of projection means and a plurality of light responsive means.

10. The system of claim 5 wherein the projection means has a housing with means therein for rotatably supporting the encoding member the encoding member being optically encoded, spaced first and second sources of light in the housing at locations on one side of the optically encoding member, and a separate lens assembly associated with the first and second light sources in position to focus light from the respective light sources onto the object.

11. The system of claim 5 wherein the means responsive include video camera means and means associated therewith to produce a raster onto which the images formed from the patterns projected onto the object by the projection means as viewed thereby can be registered.

12. The system of claim 5 wherein the positions and the widths of the light conducting and non light-conducting areas on the encoding member are established according to a multi-position code having one position for each light conducting and one position for each non light-conducting area in each segment.

13. The system of claim 12 wherein each code position is represented by a six-bit binary word.

14. The system of claim 13 including a solution code associated with the multi-position code, said solution code having a plurality of positions each of which corresponds to one or more positions in the multi-position code.

15. Means for producing a measured representation of a boundary formed by and between adjacent lighted and non-lighted areas projected onto an object comprising

a projector having a light source and a lens system positioned for projecting light from said source onto an object,

a film member positioned in alignment with the light source and the lens system, said film member having at least two adjacent distinct areas each formed of differently optically encoded patterns including patterns formed by adjacent relatively transparent and opaque areas through which light from the light source passes as it is projected by the lens system onto the object,

means to move the film member in relation to the light source whereby at least two distinct areas of encoded patterns are projected onto the object at different times,

camera means spaced from the projector and oriented to view the object and at least a portion of the encoded patterns projected thereon by the projector, said camera means having a lens system focused on the object and on an image plane in the camera means onto which the viewed pattern on the object is focused,

means for electronically producing a separate grid pattern onto which the different focused patterns of the object as seen by the camera means are applied when each of the distinct differently optically encoded patterns is projected thereon to produce image representations thereof on the respective separate grid patterns,

means for storing the separate representations produced when the at least two differently optically encoded patterns on the film member are projected onto the object,

means including first image scanning means for producing at least one pair of image coordinate measurements representative of selected locations on one of the grid patterns, said measurement pair for each location identifying the coordinates of the location on the grid pattern where a linear feature occurs as formed by and between adjacent transparent and opaque encoded areas on the film member as projected onto the object occurs, another measurement indicating whether the light transition occuring thereat as sensed by the scanning means is from a light to dark or from a dark to light transition in the projected pattern, and

means including other image scanning means for producing another bit of information from the grid pattern onto which the different optically encoded patterns are projected, said other scanning means responding to the location information produced by the first scanning means to produce information for each location for which measurement pairs are produced indicating at the corresponding locations on the grid pattern whether that location on the grid pattern on which the different optically encoded images are projected is in a light or dark portion of the pattern.

16. The means of claim 15 including means for storing information representative of the location on the respective grid patterns where at least two linear features occur, and

computer means programmed to produce a three dimensional representation of at least a portion of the object along which one of said linear features occurs, said one linear feature being produced by the projector projecting one of said optically encoded images onto the object and by the camera means viewing at least a portion of one of the linear features from its different orientation relative to the projector.

17. Means for projecting optically encoded patterns onto a remote location comprising a projector having a spaced light source and lens system and optically encoded pattern forming means positioned in optical alignment therewith, said pattern forming means including a disc having annular patterns of light encoded areas positioned thereon, means for rotating the disc so that the annular encoded areas move relative to the light source and the lens system, the annular area including a plurality of adjacent circumferentially positioned segments each formed by a plurality of adjacent light conducting and non light-conducting areas capable of projecting patterns which are characterized by having distinct boundary lines formed by and between each adjacent light conducting and non light-conducting area, the locations of the boundary lines projected by the areas in each segment being located at distinctively different positions.

18. The means of claim 17 wherein the coded light conducting and non light-conducting areas are circumferentially extended elongated areas.

19. The means of claim 18 wherein the locations of the light conducting and non light-conducting areas in the different segments are determined according to a code characterized by a base subpattern of coded information including a distinct first set of bit information for each segment, said bit information for each set corresponding to a unit width measure of displacement along the width of said segment, said sets including bits of information representative of the width of the light conducting and the non light-conducting areas and hence the locations where the projected linear features occur, the formula for the code being such that no two linear features occur at the same location.

20. The means of claim 19 wherein information in each set is such that all projected linear features are separated by a predetermined minimum distance.

21. The means of claim 20 wherein said minimum separation distance is 3 unit widths of displacement.

22. The means of claim 17 wherein the locations of light-conducting areas and non light-conducting areas on the encoded disc are determined by trial and error.

23. The means of claim 19 including a second subpattern of coded information for extending the base subpattern of coded information to create a longer pattern of coded information, said second subpattern being derived from said base subpattern.

24. The meass of claim 23 wherein said second subpattern is derived from said base subpattern by cycling and shortening said base subpattern such that the longer pattern maintains compliance with the formula code used to create the base subpattern.

25. The means of claim 24 including additional subpatterns for further extending the longer pattern, said additional subpatterns being derived from the existing base and second subpattern.

26. A system for creating data to represent a non-patterned image of a three-dimensional surface comprising

means for projecting onto the surface a sequence of patterns formed by passing light through patterned masks chosen such that selected locations on the surface are illuminated by at least one of the projected patterns in the sequence,

sensor means positioned at a spaced location from the surface in position to observe distinctive projected pattern views on the surface and to produce image representations thereof, and

means for processing image representations of at least two projected pattern views as observed by the sensor means, said processing means including means establishing a maximum intensity of the images at predetermined locations of the images being viewed containing processing data representative of the relative positions in space of the surface to be represented, the projection means and the sensor means, said processing means establishing at predetermined locations of a selected one of the image representations a maximum intensity of selected images thereat.

BACKGROUND OF THE INVENTION

The present invention resides in a novel system for obtaining information representative of the three-dimensional shape of an object in space and more particularly to a novel manner of developing projection patterns that can be projected onto the object each of which is imaged, the images then processed together in a certain way so as to measure the surface form of the object. It is especially important to the invention to be able to project a number of patterns having light and dark areas, the positioning of such areas in the different patterns being formed so as to have certain properties of the edges that are formed by and between adjacent light and dark areas in the different patterns, which correspond to projected surfaces in space, a major purpose of which certain properties is to maximize the number of such surfaces that can be projected while maintaining the ability to unambiguously identify each projected surface using only the sequence of images of the object. Profile lines or edges produced by different patterns will fall on the object to be reproduced. The pattern of the alternating light and dark areas is chosen to facilitate identification of these projected edges that fall on the object and are subsequently viewed and processed to measure the surface form of the object.

There are various systems for obtaining data useful to produce three-dimensional representations of an object. Such systems generally include a projector of radiant energy and a corresponding image recording means. The three-dimensional surface recording technology has grown substantially over the last several decades resulting in even more such systems. One such system, which like the one disclosed here and many others, dating back to early in the 20th century, e.g. Smith in U.S. Pat. No. 891,013, and Edmonds in U.S. Pat. Nos. 1,485,493, 1,615,261 and 1,716,768, and earlier by Willeme in U.S. Pat. No. 43,822, and more recently by Cruickshank, U.S. Pat. No. 4,613,234, and patent application Ser. No. 786,322, and Morioka in U.S. Pat. Nos. 3,580,758, 2,066,996, 2,350,796, 2,015,457 and 1,719,483; and in British Patent No. 439,448, as well as Jeffreys British Patent No. 471,617, comprises a camera-projector pair for the development of a data file which can be stored or used for subsequent representation of three-dimensional surface configurations, and is disclosed by DiMatteo et al in U.S. Pat. No. 3,866,052. With this and all such similar types of systems, the three-dimensional object to be recorded is placed in the field of projection of a light or other type of radiant energy projector, wherein the pattern of the projected light is structured in accord with the method of the particular invention. The surface of the object intersects with the projected light pattern in forming the reflected radiant energy. The radiant energy reflecting from the object is also within the field of view of the objective lens associated with a camera element. The geometric fixed relationship between the object, projector and camera lens is known and such information is subsequently employed together with the reflected radiant energy pattern in representating the surface configuration.

A concern of manufacturers and users of such systems is how to define a coordinate system which can be maintained and which facilitates the gathering of reliable data about the location and surface characteristics of the subject object; and from this information to identify with precision the spatial location of a point or series of points on the object's surface.

The approach claimed in U.S. Pat. No. 3,866,052 is well known in the prior art. See, for example, D. Calas, "Theory and Computer Implementation of Image Processing by Boolean Filters", Washington University of St. Louis Master's Thesis, 1970, as well as his references to preceding literature.

In light of the foregoing comments, it will be understood that a principal object of the present invention is to disclose a method of obtaining higher resolution three-dimensional representions of an object from fewer sequences of projections and recordings than is presently possible, and to obtain a given resolution with fewer projections and recordings. This also means that fewer projections and much less time is required by the present system to obtain data from which to gather information as to the shape of a surface contour. If the information is to be used in the reproduction of the shape of an object or person it means that the object or person needs to pose for a very short time, typically less than one second.

An additional object of the invention is to provide a method of digitizing data to enhance its interpretation and to produce a three dimensional representation of a surface.

Another object of the invention is to teach new and enhanced methods of profile line identification in the recorded image.

Another object is to locate, identify and develop data representative of profile edges projected onto an object and viewed by camera means.

Another object is to provide means to determine the location of points in space along a profile line projected onto an object and to produce data representative of said points in space.

A further object of the present invention is to provide a method of viewing, digitizing and processing of data to produce a quantitative measure of the viewed object.

A still further object of the present invention is to provide means of obtaining data useful to create an enhanced detailed representation of a viewed object without increasing the number of required mask segments.

Another object is to provide a three-dimensional representation means to enhance and ensure the accuracy of obtained data by systematically locating and identifying the location of profiles in a viewed image.

A further object is to obtain data corresponding to surface characteristics irrespective of the surface reflectance characteristics of the scanned object.

Yet a further object is to provide a novel method of accurately locating profiles using projected patterns and their functional inverses.

These and other objects and advantages of the present invention are realized by the present system which is based upon the light beam profiling principles described in John Cruickshank's portrait sculpture U.S. Pat. Nos. 3,796,129, 3,690,242 and 3,688,676. The basic concept disclosed in the Cruickshank patents is to recreate three-dimensional objects without requiring physical contact with the sensed object.

The method described in the Cruickshank patents utilizes the projection of a single planar surface, or sheet, light from a single projector to intersect the surface of the subject to be sensed; and a single photographic camera, positioned apart from the projector, to view and record an image of this light intersection, or "profile". In these patents it was shown that by knowing the positions and orientations of the projector and camera, as well as the focal length of the camera lens, the image of that profile can be projected onto a screen placed at a proper distance and angle from the projector to form a viewable image of a profile corresponding in size and shape to that created upon the original viewed surface. Through the process of moving the sensed surface and the projected light plane relative to one another, with or without concurrent movement of the camera as necessary, such profiles can be repeated at multiple positions over the surface so as to represent multiple profiles, which through a process of interpolation, can be used to represent the sensed surface. As disclosed in the Cruickshank patents, the system was used to manually trace the projected profiles so as to control a cutting machine to produce an approximate replica of the sensed object's surface.

The present system is an important improvement over the known prior art, including the Cruickshank patents. To generate a representation of part or all of a subject surface with the present system, spaced multiple light sheets produced by light sources including a laser light source are projected at the same time. Here, we include "light sheets" and "profiles" to be defined in the edge information between light and dark areas, that projected sheet or boundary forming a surface in space, intersecting in a line on the subject called a "profile". They intersect the subject giving it a zebra like appearance and are viewed together. With this system, instead of requiring N images to be processed for N profiles, multiple boundary surfaces are projected at once and only a smaller number of images need be viewed and processed to locate the resultant intersection profiles due to a novel technique for creating the patterns which uniquely identifies and corresponds the profiles in the viewed image with the projected surface that created them.

Once the imaged profiles are identified, if it is desired to sense more of the surface than is in common view between one projector of multiple surfaces and one camera, then the projector, subject, and camera, or any one or two of these, can be moved and the process repeated to obtain information about additional portions of the surface. The amounts of any such movement must be known quantitatively and employed in the solution. It is also contemplated to use a plurality of spaced stationary projectors and cameras, preferably in a darkroom, and to strobe the projectors in a sequence so that all or any desired surface portion of the object can be viewed and the data obtained processed to produce either a part or a full three-dimensional representation.

The projection patterns employed with the present system are uniquely designed in accordance with the present invention to facilitate identification and correspondence of the profiles in the viewed image with either the light sheets or the boundaries, or edges, between light and dark portions of the projected pattern. Briefly, the scheme is one of arranging light and dark banded areas on at least two projected patterns, such that they meet certain properties. One method of generating a set of patterns that meet these properties is described. In this example, each pattern correlates to a 16-bit long cyclic generatric code. Using this code as a basis, a much longer code is generated that has the desired properties of uniqueness throughout its length, and meets the chosen properties through out its length. By locating a profile in the viewed image, through a process of reading the code correspondng to the viewed image, the corresponding profile in the projected pattern can be identified, and the accuracy of the identification can be checked and if desired cross-checked.

The patterns also may be projected in the form of thin sheets of light rather than dark-to-light or light-to-dark boundaries or edges. This is done by projecting a pattern of thin light lines on a dark background, or vice versa.

The more boundaries or light sheets that are projected at once, the truer the resultant surface modeling. However, if they become too closely spaced, then optical and camera resolution may not separate them in the viewed image, thus losing the information carried by the denser profiles.

The exact spacing is also a subject of the present invention and is detailed in what follows herein. An important ad