WO1994003868A1

WO1994003868A1 - Method of and apparatus for generating three dimensional models

Info

Publication number: WO1994003868A1
Application number: PCT/GB1992/001436
Authority: WO
Inventors: Michael Ian Burman; Mark Nicholas Forrest; Peter James Langley
Original assignee: Deantree Ltd.
Priority date: 1992-08-03
Filing date: 1992-08-03
Publication date: 1994-02-17

Abstract

Creating scale models of complex three dimensional structures such as oil bearing rock strata has hitherto been achieved using costly and time consuming conventional model making techniques relying on materials such as 'papier mache' and plastics. This invention shows a computer system that operates on data generated by conventional 3D CAD software techniques to extract the contours of the object to be imaged at a number of user defined heights. The area bounded by the contours for each elevation is then subject to a 'fill' painting process and the resultant filled contour then printed onto a sheet of acetate. When the acetates for all the filled contours are assembled as a stack, a three dimensional image of the structure is formed.

Description

Method of and apparatus for generating three dimensional models

Field of the Invention

This invention relates to a method of and apparatus for generating three dimensional models, and in particular to such a method and apparatus that enables the production of three dimensional models of complex geographical topologies such as oil bearing rock strata, multi-layered structures such as micro-processors, and complex volumes such as organs of the human body.

Description of the Prior Art

There are a number of different approaches to three dimensional modelling. One traditional method involves creating a scale model using papier mache, plastic foam or some other such modelling material. This approach is widely used for terrain modelling, applicable for example when designing a golf course, or for depicting the complex strata and fault lines of oil-bearing rocks. One advantage which scale models possess is that they are permanent and generally easily understood by non-specialists. They are, however, expensive and time consuming to design and manufacture since they generally involve a skilled model maker following detailed plans or records to laboriously build up the model.

An alternative to such traditional model making is to generate a computer model for display on a conventional display device such as a cathode ray tube [CRT], liquid crystal display [LCD] or, where a permanent record is needed, a pen plotter. This approach does not result in a three dimensional model as such, but instead merely a two dimensional image of a three dimensional model shown on the CRT or pen plot. The computer model of a terrain, for example, will itself typically consist of a mathematical description of the terrain, comprising a vector database of the exact co-ordinates of features of significance, such as roads or areas of interest. The model can allow analysis of diverse relationships between features and ensure that the location of new features, such as bore holes, can be very accurately added to the existing model. The terrain can be illustrated on the display from a variety of perspectives and stand points. AutoCAD® from Autodesk Inc of 2320 Marinship Way, Sausalito, California 94965, USA is a widely used software modelling program that allows such 3D models to be built up using a variety of different approaches. One approach is to define the entire surface of the model in terms of small polygons. Another approach, requiring the Advanced Modelling Extension to AutoCAD®, allows models to be constructed that the computer treats as solid, thereby allowing various solid object manipulations to be performed. Typically, such a program can be used to generate permanent plans by providing output to conventional printing devices such as a pen plotter or thermal transfer plotter.

Another field that utilises sophisticated computer imaging of three dimensional objects is that of medical imaging, e.g. magnetic resonance imaging [MRI]. Typically, in MRI, data defining transverse slices through the organ being imaged is obtained. Each slice can then be separately displayed in plan view on a CRT. Various manipulations can be performed on the data, for instance to generate an image of the organ reconstructed as a series of parallel slices.

Generally, with all such computer based systems, the output is merely a two dimensional representation of a three dimensional object, and is therefore less useful for many purposes than a three dimensional scale model. However, some advantages that computer generated output has over more traditionally crafted models are speed, accuracy and the facility to display complex relationships between regions.

Further reference may also be made to the standard laboratory microscopy technique of slicing an object into a series of thin, translucent layers, each layer being viewable using a conventional light microscope. Objects of the Invention

It is an object of the invention to provide a method of and apparatus for generating three dimensional images that combines the speed, accuracy and flexibility of computer generated modelling with the comprehensibility and impact of scale modelling.

Statement of the Invention

In accordance with a first aspect of the present invention, a method of and apparatus for generating a three dimensional image of a subject involves using a computer system that operates on data, generated by conventional 3D CAD software techniques, to extract the contours of the subject to be imaged at a number of user defined elevations. The area bounded by the contours for each elevation is then subject to a 'fill' painting process and the resultant filled contour then printed onto a sheet of acetate. When the acetates for all the filled contours are assembled as a stack, a three dimensional image of the structure is formed. This approach allows realistic three dimensional models of complex structures to be created far more quickly and cheaply than with conventional model making techniques. Additionally, because the resultant display is actually a three dimensional model, it is more striking and easily understood compared to a computer simulation on a flat display screen.

In a more general aspect, the invention relates to a method of generating a three dimensional image of a subject comprising the steps of: defining portions of the subject in terms of a series of data groups, each data group defining an individual such portion; providing a series of substantially translucent layers; forming an image derived from each data group onto a layer, to provide on that layer a representation of a said portion of the subject; building up a series of such representations by forming each of the images onto a layer to form a series of layers, the images being derived from data groups defining successive, adjacent portions within the subject; arranging the series of layers to display the three dimensional image of the subject that consists of an assembly of representations of successive, adjacent portions of the subject.

In a still further aspect of the invention, there is provided an apparatus for generating a three dimensional image of a subject which comprises: means for defining portions of the subject in terms of a series of data groups, each data group defining an individual such portion; means for forming an image derived from each data group onto a substantially translucent layer to provide on that layer a representation of a said portion of the subject; means for building up a series of such representations by forming each of the images onto a layer to form a series of layers, the images being derived from data groups defining successive, adjacent portions within the subject; means for arranging the series of layers to display the three dimensional image of the subject that consists of an assembly of representations of successive, adjacent portions of the subject.

Further, in either aspect, the portions may be the boundaries of sections through the subject, or sections through the subject itself.

Additionally, the data group may be generated by merely recording an image of the portion using a camera to describe the portions.

Preferably, however, the data group is generated or read by a CAD program. More particularly then, the following will also be provided for: utilising a three dimensional modelling program on a computer to generate a model of the subject; deriving the data groups from the said model; generating output data which defines an image derived from each data group; using the output data to control a printing device to produce the said representations. In such an embodiment, the three dimensional modelling program preferably generates a three dimensional model of the subject defined as an ASCII file or as three dimensional lines and polylines. The data set may then comprise a three dimensional triangulated irregular network [TIN] corresponding to the model from which may be subsequently generated a set of contours defining the network at user determined elevations from a specified axis.

In another embodiment, the shading of one feature is produced by ink or dye which fluoresces within a first light frequency range and the shading of a further feature is produced by ink or dye which fluoresces within a second light frequency range. Successively illuminating the representations with light in the first and second frequency ranges will then allow a variety of different visual effects to be achieved.

Preferably, each layer is manufactured from acrylic sheet and the image is silk screened onto the acrylic sheet.

Alternatively, each layer is manufactured from acetate, preferably premier tri-acetate [PTA] and is positionable between transparent acrylic interleaves, of preferably matching refractive index to the PTA, for secure locking in mutual alignment so that the layers can be arranged as a stack to display a three dimensional image of the subject.

The acrylic interleaves are mountable into a cartridge that can fit over a light box, illuminating the bottom of the stack so that the image appears to float within the stack. Conveniently, the cartridge is removably positionable over the light box and the light box is rotatably mounted on a yoke that allows the light box and cartridge to be rotated so that the image can be viewed from a variety of angles.

In a still further embodiment, the substantially translucent layer is a liquid crystal display (LCD) panel, such as used in LCD projection systems. The LCD panel therefore replaces a layer, such as a PTA layer. The panel can be controlled to display the image that would be shown on a PTA layer by conventional techniques; it is more flexible than the PTA approach, however, since the image can be readily altered. In practice, a stack of LCD panels may be placed on top of one another to provide a three dimensional LCD device.

Brief Description of the Drawings

A method of and apparatus for generating three dimensional models will be further described with reference to the following figures of which:

Figure 1 is a perspective view of a three dimensional image of an apple generated by the stacking of glass layers to which photographic images of individual slices have been applied;

Figure 2 is a perspective view of a three dimensional contour map of rock formations generated using CAD software, the map comprising a stack of acetate layers on which a contour has been plotted using a conventional thermal transfer plotter. Figure 3 is an exploded view of the stack of acetate layers of Figure 2 showing the registration pegs which secure the layers in accurate mutual alignment.

Figure 4 is a schematic cross sectional view of the cartridge comprising acetate and acrylic layers in a ring binder.

Figure 5 is a schematic cross sectional view of the combination of light box and cartridge.

Detailed Description

In its simplest embodiment, the invention involves obtaining a plan view image of sections through the subject to be modelled and then applying those images onto a series of transparent layers to reconstruct the sections. For example, where an apple is to be modelled, as illustrated in Figure 1, it is first cut into a series of transverse slices and a photograph taken of each slice to provide a plan view image of that slice, each photograph constituting one of the data groups referred to previously.

Each image is then applied to a thin transparent sheet 1. This can readily be achieved by enlarging the photographic negative onto a sheet 1 coated with light sensitive emulsion. Such sheet materials 1 are widely available and will not be further described here. Each sheet 1 is then arranged in a stack 3 by placing the sheets 1, in appropriate order, over registration pegs 4 that pass through holes in each corner of each sheet 1. Consequently, a model of the apple 2 is constructed from the stack of sectional images. It will be apparent that the distance between each sheet 1 should correspond to the distance between the associated slices of the original apple. This is achieved in the illustrated embodiment by mounting spacers [not shown] between each thin sheet 1. Alternatively, each sheet 1 could be mounted in a cardboard frame of required thickness.

In another embodiment, the invention may be used to provide output for a CAD system, for example a Geographic Information System [GIS]. GIS is computer based and comprises a computerised data base typically defining geographic, descriptive and statistical data. The data is obtained using conventional techniques- for example, remote observation by satellite, seismic probing etc. and is input to the computer programmed with GIS in a conventional manner. GIS allow the analysis of complex relationships between data in the database and is widely used by civil engineers, architects, oil exploration geologists and cartographers. AutoCAD® provides a GIS capability which allows the subject to be modelled; for example, a volume of complex rock strata can be defined in terms of three dimensional lines and polylines derived from digitised contours or defined break lines defining the strata.

In a preferred embodiment, the three dimensional data from AutoCAD® is read by a digital terrain modelling program such as AutoCivil, obtainable from 4 The Causeway, Sutton SM2 5RS, Surrey, UK. AutoCivil generates a digital terrain model from the data in the form of a three dimensional TIN [triangulated irregular network] and stores the TIN to disk. AutoCivil then generates contours between user specified levels and intervals.

For example, with the subject illustrated in Figure 2, seismic data relating to oil bearing rock formations was collected and stored on a digital computer. The data defines the various important features that geologists require in order to understand the region, including fault lines, strata etc. The exact nature of the data is conventional and will not be further described here. The data was processed by AutoCAD's digital contouring feature, and the resultant three dimensional lines and polylines processed by AutoCivil into a TIN. The next step is for the user to select the desired positions of each slice through the TIN, i.e. the depth of each transverse slice through the model.

For the avoidance of doubt, it is noted that the present invention is not limited to transverse slices but includes any slice or section, including non-planar sections. It is further noted that certain structures are more clearly modelled by introducing distortions into the model. For example, certain regions of the rock strata may have particularly interesting features and the user will require that more detail is provided in those regions, whilst still indicating how those regions fit into the general picture. This may be achieved by increasing the distance scale at the region of interest.

After the user has selected the required slice positions, AutoCivil presents the user, on a CRT display connected to the computer running AutoCivil, with an image of the contour for each slice through the TIN so that the user can verify that the contour is closed, or.runs to the edge of the model. This step is necessary since at a later stage, the area bounded by the contour will be filled with a selected fill pattern by a conventional paint program. If the contour is not closed or connects with the edge of the model, then the entire screen area will be filled with the fill pattern. For contours that are not closed, AutoCAD can be used to effect proper closure. Once the user has verified that all contours are closed, AutoCivil extracts the graphical information and passes it onto a paint program such as Dr Halo. The paint program effects the fill as described above.

In a preferred implementation, a contour in any given slice is filled only in those areas not overlaid by a shaded part of the contour of the overlying slice. This can be effected by subtracting the overlying area from the immediately underlying area and only filling the remaining area. This may be achieved by conventional subtraction techniques that will not be detailed here. This approach is useful when depicting land forms since if the whole bounded area of each contour progressing down through the model were shaded, then the resulting depiction would be confusing. Additionally, this approach allows features that are inside other features to be picked out: for example, the line of a vertical bore hole can be clearly picked out as it passes through the apex of a conical rock formation and down inside the rock formation since the inside of the formation is not shaded at all - only the parts of the conic that extend out beyond the overlying rock are shaded.

Once appropriate filling has been completed, the image of each slice is output to a thermal transfer plotter, typically a Tektronix® Phaser III PXI. The plotter prints an image of the filled contour 6 in colour onto an A3 sheet of premier tri-acetate [PTA] film, of 100 microns thickness. It is desirable to add an anti-static agent to the PTA film since any static build up can rapidly attract dust, which inhibits the adhesion of the coloured wax from the plotter. PTA has been found to be particularly suitable for the present application since it is optically very clear, allowing up to 21 sheets to be layered on top of one another in constructing the final stack for viewing. Additionally, it is stable up to approximately 140° C, making it compatible with the Tektronix thermal plotter that operates within the 38° C to 58° C range. In some cases, it may be preferable for each PTA sheet to have a backing sheet to aid handling of the sheet through the printer and improve print quality. PTA is obtainable from Film Sales of 145 Nathan Way, Woolwich Industrial Estate, London, SE 28 OBE, England.

As an alternative to PTA, a variety of other films may also be found to be suitable. For example, polyester film from Imperial Graphics Products Ltd of Imperial Works, Weaste, Salford, M6 5RH, England is widely used for many different thermal printers and may be suitable for certain embodiments of the present invention, particularly those requiring fewer than 20 sheets stacked on top of one another.

Referring now to the Figures 2 to 5, each PTA sheet 5, including the appropriately shaded contour 6, is then loaded into the viewing cartridge, indicated generally at 10, by locating registration holes [not shown] in each sheet 5 over registration pins 1 1 and then lowering the sheet 5 over the pin 11.

The cartridge 10 comprises a number of acrylic sheets 12, typically made of Perspex®, each being held in position by a pair of ring binders 13. For the avoidance of doubt, the illustrations are schematic; particularly, the sheets 12 are not drawn to scale - in practice, there would be up to 21 PTA sheets interleaved between acrylic sheets.

In use, a Perspex® sheet 12 is positioned over each PTA sheet 5 so that the PTA 5 is held flat. Consequently, a stack of PTA sheets results. Additionally, the width of the Perspex® sheet 12 can be such that the separation between adjacent PTA sheets 5 is in scale to the subject that is represented. Alternatively, spacers or additional perspex sheets [not shown] can be placed between adjacent Perspex® sheets, on which PTA sheets lie. The Perspex® is preferably ICI® OX002 clear grade, of thickness typically in the range of 2mm to 5mm, although even thicker gauges are possible if the model demands it. The refractive index of the Perspex® matches that of PTA; where material other than PTA is used, it too must have a refractive index that matches the Perspex® interleaves.

There is provided a lid 18, having a large transparent window 17 for its top surface, but opaque side walls. Shielding the edges of the PTA and Perspex® sheets from view adds to the illusion of the three dimensional image floating in space. The stack of Perspex® sheets 12 and PTA sheets 5 are locked together by a simple cam arrangement [not shown], ensuring that mutual alignment is preserved even when the cartridge 10 is moved. Conveniently, the cartridge 10 can then be dropped onto the light box unit 14. The light box 14 comprises low heat emission lamps 15 such as the Wotan PL18, positioned under a diffusing sheet 16. In this way, the stack of PTA sheets 8 is illuminated from below, allowing a viewer to see the reconstructed image through the window 17 of the lid 18 of the cartridge. The light box is mounted on a stand 20 and is counterbalanced by a yoke so that it can be spun around on an axis 21 to enable the model to be viewed from different angles. When the lights 15 are turned on, the three dimensional reconstruction of the subject, formed from the stack of PTA slice images, appears to float within the stack 8.

The above described embodiment is particularly appropriate where the user requires a rapid output. In many instances,' there is no such requirement. Then, the user may obtain a higher quality set of images, not limited to the 21 layers of PTA, by arranging for the image of each slice to be silk- screened onto the Perspex® sheets. Because Perspex® is optically clearer than PTA or polyester, more than 21 layers can be built up without loss of image quality. Further, the silk-screening process can give slightly higher image quality than a thermal plotter. The process of silk-screening comprises the steps of generating the PTA sheets as before, with the modification that the shading used in the fill process should be black. The PTA sheet can then be photographed and the negative used as in a conventional silk-screening process to impart an ink image onto the Perspex® sheet. The process is repeated for each slice and the set of Perspex® sheets loaded into the cartridge system described above for final viewing. This embodiment is suitable for producing permanent displays, such as foyer displays or museum plans.

In a further embodiment, not illustrated, the shading of features of interest is carried out using UV dyes, i.e. dyes that are visible to the eye only when illuminate by UV light of a certain frequency. The advantage of this approach is that the model can incorporate details of structures that are revealed only when the UV light is turned on. Since there are a variety of different UV dyes, sensitive to different frequencies, it is possible to conceal a number of different features. By arranging for the illumination to successively pass through the frequency range that activates the different dyes, it is possible to produce animation of features of the model. In a still further embodiment, not illustrated, the model can include shading of one feature in one colour, and shading of another feature in the complimentary of that colour. When the light which illuminates the model is of the first colour, achieved by placing a filter of that colour over the light sources, then the second feature appears very dark, with the first feature virtually invisible. Likewise, when light of the second colour is used to illuminate the model, it is the first feature that appears very dark, with the second feature virtually invisible.

Claims

1. A method of generating a three dimensional image of a subject comprising the steps of: defining portions of the subject in terms of a series of data groups, each data group defining an individual such portion; providing a series of substantially translucent layers; forming an image derived from each data group onto a layer, to provide on that layer a representation of a said portion of the subject; building up a series of such representations by forming each of the images onto a layer to form a series of layers, the images being derived from data groups defining successive, adjacent portions within the subject; arranging the series of layers to display the three dimensional image of the subject that consists of an assembly of representations of successive, adjacent portions of 'the subject.

2. The method of Claim 1 wherein the portions are the boundaries of sections through the subject.

The method of Claim 1 wherein the portions are sections through the subject.

4. The method as claimed in Claim 1 comprising the further steps of: utilising a three dimensional modelling program on a computer to generate a model of the subject; deriving the data groups from the said model; generating output data which defines an image derived from each data group; using the output data to control a printing device to produce the said representations.

5. The method of Claim 4 wherein the step of utilising a three dimensional modelling program generates a three dimensional model of the subject defined as an ASCII file or as. three dimensional lines and polylines.

6. The method of Claim 5 wherein the step of deriving the data groups comprises the further step of generating a three dimensional triangulated irregular network corresponding to the model and subsequently generating a set of contours defining the network at user determined elevations from a specified plane.

7. The method of any preceding claim comprising the further step of arranging for the representations to include shading or colouring to represent a feature of the subject.

8. Trie method of Claim 7 as dependent on Claim 6 wherein the area bounded by a contour at a given elevation is the said feature and the shading is performed by a paint fill procedure.

9. The method of Claim 8 wherein a feature that is represented in successive adjacent layers is shaded in the uppermost representation, and is shaded in lower representations only in those areas not overlaid by a shaded area.

10. The method of Claims 7 - 9 wherein the shading of one feature is produced by ink or dye which fluoresces within a first light frequency range and the shading of a further feature is produced by ink or dye which fluoresces within a second light frequency range, at least one of the ranges not being in the visible spectrum.

11. The method of Claim 10 comprising the further step of successively illuminating the representations with light in the first and second frequency ranges.

12. The method of any preceding claim wherein each layer is manufactured from acrylic and the step of applying the image comprises the step of silk screening the image onto the acrylic.

13. The method of any preceding Claims 1 - 11 wherein each layer is manufactured from acetate.

14. The method of Claim 13 wherein the acetate is premier tri-acetate.

15. The method of Claims 13 or 14 comprising the further step of adding an anti¬ static agent to the acetate layer to prevent the layer attracting dust.

16. The method of Claims 13-15 wherein each acetate layer is placed between transparent acrylic interleaves of matching refractive index and secured in mutual alignment.

17. Apparatus for generating a three dimensional image of a subject comprising: means for defining portions of the subject in terms of a series of data groups, each data group defining an individual such portion; means for forming an image derived from each data group onto a substantially translucent layer to provide on that layer a representation of a said portion of the subject; means for building up a series of such representations by forming each of the images onto a layer to form a series of layers, the images being derived from data groups defining successive, adjacent portions within the subject; means for arranging the series of layers to display the three dimensional image of the subject that consists of an assembly of representations of successive, adjacent portions of the subject.

18. The apparatus of Claim 17 wherein the portions are the boundaries of sections through the subject.

19. The apparatus of Claim 17 wherein the portions are sections through the subject.

20. The apparatus as claimed in Claim 17 wherein a three dimensional modelling program on a computer is used to generate a model of the subject and derive the data group from the said model, further comprising; means for generating output data which defines an image derived from each data group; a printing device using the output data to produce the said representations.

21. The apparatus of Claim 20 wherein the three dimensional modelling program generates a three dimensional model of the subject defined as an ASCII file or as three dimensional lines and polylines.

22. The apparatus of Claim 21 wherein the three dimensional modelling program generates a three dimensional triangulated irregular network corresponding to the model and subsequently generates a set of contours defining the network at user determined elevations from a specified plane.

23. The apparatus of Claims 17 - 22 comprising program means for arranging for representations to include shading to represent a feature of the subject.

24. The apparatus of Claim 23 wherein the program means is arranged so that a feature that is represented in adjacent layers is shaded in the uppermost representation, and is shaded in lower representations only in those areas not overlaid by a shaded area.

25. The apparatus of Claim 24 wherein the area bounded by a contour at a given elevation is the said feature and the shading is performed by a paint fill procedure.

_* 26. The apparatus of Claim 23 - 25 wherein the shading of one feature is produced by ink or dye which fluoresces within a first light frequency range and the shading of a further feature is produced by ink or dye which fluoresces within a second light frequency range, at least one range not being in the visible spectrum.

27. The apparatus of Claim 26 comprising means for illuminating the representations with light in the first and second frequency ranges.

28. The apparatus of Claims 17 - 27 wherein each layer is manufactured from acrylic and the means for applying the image comprises means for silk screening the image onto the acrylic.

29. The apparatus of Claims 17 - 27 wherein each layer is manufactured from acetate.

30. The apparatus of Claim 29 wherein the acetate is premier tri-acetate.

31. The apparatus of Claim 29 or 30 further comprising a series of transparent acrylic interleaves, an acetate layer being positionable in predetermined accurate alignment between a pair of acrylic interleave so that when the acetate layers are arranged to display the three dimensional image the acrylic layers are arranged on top of one another to form a stack of acrylic interleaves.

32. The apparatus of Claim 31 wherein the acrylic interleaves are mountable onto a light box.

33. The apparatus of Claim 32 wherein the light box comprises at least one light source positioned below the stack of acrylic interleaves.

34: The apparatus of Claim 33 wherein the stack is removably positionable in the light box and the light box is rotatably mounted on a yoke that allows the light box and stack to be rotated so that the image can be viewed from a variety of angles.