WO1993004437A1

WO1993004437A1 - Efficient storage of geographic data for visual displays

Info

Publication number: WO1993004437A1
Application number: PCT/US1992/005175
Authority: WO
Inventors: Mark H. Johnson
Original assignee: Hughes Aircraft Company
Priority date: 1991-08-15
Filing date: 1992-06-22
Publication date: 1993-03-04

Abstract

A geographic image processing system (10) includes an elevational data unit (12), an imagery data unit (14), an elevational data processor (16), an image generation processor (20) and a display unit (22). The elevational data processor (16) greatly reduces the number of elevational data points by identifying hills and valleys in the elevational data. The hills and valleys are identified by identifying points (26) having significant differences in slope along an elevational curve (24) in said elevational data and then creating tiles (30) and lines (28). By storing and processing only information about the tiles (30) and lines (28), the amount of elevational data required to produce a three dimensional image is greatly reduced. This results in much faster throughput for the image generation system (20).

Description

EFFICIENT STORAGE OF GEOGRAPHIC DATA FOR VISUAL DISPLAYS

BACKGROUND OF THE INVENTION

1. Techni cal Fi el d

This invention relates to a system for processing geographic data, and in particular to a system for efficiently storing and processing elevational geographic data.

2. Discussion

Geographic information such as image and elevation data is employed in many applications. The combination of elevation and image data can provide relatively realistic simulations of a geographic region. This information may be used, for example, in simulation systems where it is desirable to simulate geographic terrain in three dimensions. Also, such simulations are useful as a temporary radar substitute to permit an aircraft to determine the location of nearby geographic features such as hills, while flying at low elevation. Such image and elevational data is often gathered by using satellite imagery. In one technique, images of a geographic region are taken by the satellite from different viewing angles. By analysis of the same region from different perspectives, the elevation of particular points can be determined by measuring the apparent shift of features from the different views. For example, this kind of elevational data is available from the Defense Mapping Agency and is called -Digital Terrain Elevational Data" (DTED) .

Such data is typically stored on a grid with elevation values at each coordinate location on the grid. Depending on the image data, the resolution of the elevation data may be better or worse than the pixel imagery information. In any event, because there are many elevation data points (one at each coordinate location), the process of producing and displaying three dimensional images can be very time consuming. For example, these images may take the form of a computer generated perspective view of the geographic terrain. However, the excessive time required to produce an image (due to the large number of data points) limits the usefulness of these kinds of three dimensional imaging systems, particularly for real time or near real time applications, One prior approach for dealing with large quantities of elevation data has been to employ a "level of detail" algorithm. In a "level of detail" algorithm, the elevation data is reduced by using a regular subset of the information. For example, if elevation data is recorded at 100 foot intervals, the data can be reduced to 1/9 the size by using 300 foot intervals. This approach results in improved data storage for the elevation and image data but the data reduction discards significant information which will cause some hills to be ■^■squashed".

Thus, it would be desirable to provide an elevational data processing system which reduces the quantity of elevational data to thereby speed up the production of three dimensional geographic images. Further, it would be desirable to provide such a system which can accomplish this data reduction without significant loss of details and without causing hills to be squashed.

SUMMARY OF THE INVENTION

Pursuant to the present invention, a system and method is provided for generating an elevational map of the physical terrain of a region. Initially, an elevational map is produced which includes an elevational value for each defined coordinate location in a coordinate system. A set of curves of the elevational value is created along a single axis of the coordinate system. Points are then identified along each of the curves which have a slope which differ from neighboring points by a value exceeding a predetermined threshold. A new elevational map containing only the identified points is then produced. Tiles in the new elevational map representing elevational features are generated by connecting each of the identified points to adjacent points in adjacent curves which are near the identified points. Finally, a three dimensional image of the region is produced from the new elevational map based on the tiles.

BRIEF DESCRIPTION OF THE DRAWINGS The various advantages of the present invention will become apparent to one skilled in the art by reading the following specification and by reference to the following drawings in which: FIG. 1 is a block diagram of the overall elevational data processing system in accordance with the present invention;

FIG. 2 is a representati e view of a geographic terrain showing mountains and valleys;

FIG. 3 is a view of the storage of elevational data for the terrain in Fig. 2 in accordance with the prior art;

FIG. 4 is a profile view of the elevational data in Fig. 3 along a single axis;

FIG. 5. is a view of the curve shown in Fig. 4 with significant points identified in accordance with the techniques of the present invention;

FIG. 6 is a representation of the geographic terrain shown in Fig. 2 produced in accordance with the present invention; and

FIG. 7 is a pair of graphs illustrating a difference technique used in the present invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring to FIG. 1, an overall block diagram of a geographic image processing system 10 in accordance with the present invention is shown. The processing system 10 accepts elevational data and imagery data, compresses the elevational data into a representation of elevational features such as hills and valleys and then combines this information with imagery information of the same scene to generate a three dimensional display of a geographic region. In more detail, elevational data is generated and/or stored in elevational data unit 12. In the preferred embodiment, the elevational data consists of geographic data structured on a grid such as the digital terrain elevational data (DTED) available from The Defense Mapping Agency. This data may comprise data representing elevation at individual coordinate points. For example, in FIG. 2, there is shown a view of the Los Angeles basis with prominent features such as the San Gabriel Mountains, fans at the base of streams, and smaller hills and features shown.

The DTED data for this region consists of a grid of elevational points, such as those shown in FIG. 3. At the DTED level one, the grid points are spaced 300 feet apart; and thus for the 60 mile x 60 mile area shown in FIG. 2, there are 1200 x 1200 points or 1.44 million points. Of course, the DTED data is available at other levels so that grid points may alternatively be spaced at 3 feet, 30 feet, and 3000 feet for example. The elevational data is received and stored in the elevational data generation and storage block 12 in FIG. 1. Also, imagery data for the same region is received and stored in the image data unit 14 in FIG. 1. This data may consist of conventional pixel information which may have greater, less, or the same resolution as the elevational data. An elevational data processor 16, shown in FIG. 1 receives the elevational data and processes this information in accordance with the present invention as described below. As a result of this processing, the elevational data processor 16 produces a compressed representation of the elevational data, containing which tiles which represent elevational features. This tile information is then transferred to an image generator unit 20 which also receives the imagery information from the image data unit 14. The image generator 20 then combines the imagery information with the compressed elevation information to produce a simulated three dimensional representation of the geographic region shown in FIG. 2. Image generator 20 may generate for example, a three dimensional perspective view from a desired perspective viewpoint. Image generator 20 may consist of a computer image generation system such as "Photovision 4», manufactured by Hughes Training Incorporated of Arlington, Texas. Finally, a display unit 22 receives the three dimensional display information and displays the information to the viewer or other control system depending on the particular application.

It can be appreciated that, with 1.44 million elevational points for example, it will be.computationally expensive for the image generation unit 20 to process all this data. In accordance with the present invention, this volume of data is reduced by identifying features representing hills and valleys in the elevational data as follows.

Referring now to FIG. 4, there is shown a profile view of the elevational data shown in FIG. 3 along a single axis. In particular, FIG. 4 contains a curve 24 which is a^" plot of elevation along the vertical axis verses geographic position in the geographic region in FIG. 2. That is, curve 24 represents the elevational data for a single vertical line along the DETD grid of elevational data shown in FIG. 3. Curve 24 indicates a peak on the left side where the San Gabriel mountains exist.

Referring now to FIG. 5, curve 24 is shown with particular points 26 identified. These points represent elevational data points on the curve 24 where the curve undergoes a relatively sharp transition. It will be appreciated that various mathematical techniques may be employed to identify points 26. In accordance with the preferred embodiment, the slope for each point may be calculated and then points identified where the difference in slope between successive points exceeds a predetermined threshold. This may be done by using conventional difference techniques. For example, referring to FIG. 7, at three points on a line (X₀, Y₀, Z₀) , (X₀, Y₀ + dy, Z.), (^χ ₀» ^Y _o ⁺ 2dy. Z_o), the slope along the Y-axis is 1. - Z_o/dy for the first two points and Z₂ - Z^dy for the last two points. The difference in slopes is Z₂ - I. - Z₀/dy which is compared to the threshold. Points 26 are then identified in this manner for each vertical data column on the grid shown in FIG. 3. Next, identified points 26 are connected to points in adjacent curves 24 that are in close proximity. For example, this may be done by connecting each identified point 26 with identified points in neighboring curves 24 that lie within a second predetermined threshold distance away from the first point 26. Once this is accomplished, the identified elevational points 26 and their connecting lines may be plotted as shown in FIG. 6. The result is a series of lines 28 and tile^'s 30 which represent elevational features such as hills and valleys in the geographic terrain.

By comparing FIG. 6 with FIG. 2, it can be seen that the tiles 30 and lines 28 roughly correspond with the hills and valleys of the geographic region shown in FIG. 2. It is important to note that FIG. 6 retains only the elevational data of these significant points 26. All other points on the curve 24 having been eliminated. This data reduction performed by the elevational data processor 16 in FIG. 1 greatly reduces the amount of data which must be handled by the image generator unit 20 in FIG. 1. It should be noted that the tiles 30 may connect adjacent identified points 26 which vary in elevation. That is, the line forming tiles may represent a sloping feature of varying elevation as well as a peak or valley of a constant elevation. This is because the connection of adjacent identified points 26 is made on the basis of their distance from each other on the coordinate grid of FIG. 3 and not on the basis of the difference in actual elevational data stored for those points.

It will be appreciated that once tiles 30 and line 28 are drawn, as shown in FIG. 6, the data may be further reduced by eliminating even identified points 26 which lie between the endpoints of straight lines. That is, only the endpoints of a straight line need be retained and points in between need not be retained. Of course, in this context the term "Straight lines- refers to lines in three dimensional space. To illustrate the magnitude of the data reduction achieved by the present invention, it is estimated that a typical mountain range in an area such as the one shown in FIG. 2 may be represented by 1000 tiles where each tile is represented by four points. Smaller hills which contain 300 -features" requiring ten tiles per feature will result in 3000 tiles. Each of these tiles may be represented by four data points. Thus, to represent a mountain range and as well as smaller hills will require 4000 tiles, or 16,000 points. This is approximately one percent of the total number of elevational points in the DTED data that previously had to be processed.

The improvements in processing speed for the reduced data may be estimated as follows. In one typical system where there are 100 pixels per tile and 36,000 tiles, it takes about 2.5 seconds to process a total image. This is calculated as follows: 2 microseconds per row setup (Step 1)

0.5 microseconds per pixel (Step 1)

4 microseconds per row setup (Step 2)

0.5 microseconds per pixel (Step 2)

10 rows of 10 pixels each [without invention] 10 rows X (2 μsec/row+10 pixel/row X 0.5 μsec/pixel)=25 μsec.

10 rows X (4 μsec/row+10 pixel/row X 0.5 μsec/pixel X 0.5 μsec/pixel) = 45 μsec.

Total/tile = 70 μsec. X 36000 = 2.52 sec. 10 rows of 100 pixels each [with invention] 10 rows X (2μsec./row + 100 pixel/row X 0.5 μsec/pixel) = 70 μsec. 10 rows X (4/_sec/row + 100 pixel/row X 0.5 μsec/pixel) = 90 μsec. Total/tile = 160 _/.sec. X 3600 = 0.576 sec.

This is a four to five times performance improvement in the processing time required by the image generation unit 20 to produce the three dimensional image.

From the foregoing, it can be appreciated that the present invention provides an elevational data processing system which greatly reduces the quantity of elevational data to thereby speed up the production of three dimensional geographic images. Those skilled in the art can appreciate that other advantages can be obtained from the use of this invention and that modification may be made without departing from the true spirit of the invention after studying the specification, drawings and following claims.

Claims

CLAIMS What is claimed is:

1. A method of generating a three dimensional image based on elevational information of the physical terrain of a region, said method comprising: providing an elevational map of said region, said map including an elevational value for each defined coordinate location in a coordinate system; creating a set of curves of said elevational values, each curve plotting the elevational values along a single axis of said coordinate system; identifying points along each of said curves which have a slope that differs from a neighboring point by a value exceed"*".; a predetermined threshold; producing a new elevational map containing only said identified points; forming tiles in said new elevational map representing elevational features in said region by connecting each said identified point to adjacent identified points in adjacent curves which are less than a second predetermined threshold distance away from each said identified point; and producing a three dimensional image of said region from said new elevational map based on said tiles.

2. The method of Claim 1 wherein said step of providing an elevational map of said region further comprises the steps of: generating multiple visible images of said region from different viewing angles; measuring the amount of the apparent shift in features at said coordinate locations between said different views; and determining the elevation at said coordinate locations based on said apparent shift.

3. The method of Claim 1 wherein said coordinate system is a cartesian coordinate system.

4. The method of claim 1 wherein said tiles are defined by at least three identified points wherein lines connecting said points form a polygon, whereby said tiles represent elevational features.

5. The method of Claim 4 further comprising the step of forming lines between neighboring identified points which do not form tiles, whereby said lines represent elevational features.

6. A method of compression of elevational data, .said method comprising the steps of: providing an elevational map of a region said map including an elevation value for each defined coordinate location in a coordinate system; creating a set of curves of said elevational values, each curve plotting the elevational values along a single axis of said coordinate systems; identifying points along each of said curves which have a slope that differs from a neighboring point by a value exceeding a predetermined threshold; producing a new elevational map containing only said identified points; forming tiles in said new elevational map representing elevational features in said region by connecting each said identified point to adjacent identified points in adjacent curves which are less than a second predetermined threshold distance away from said identified points; and eliminating all elevational data except said identified points.

7. The method of Claim 6 wherein said coordinate system is a cartesian coordinate system.

8. The method of claim 6 wherein said tiles are defined by at least three identified points wherein lines connecting said points form a polygon, whereby said tiles represent elevational features.

9. The method of Claim 8 further comprising the step of forming lines between neighboring identified points which do not form tiles, whereby said lines represent elevational features.

10. A system for producing a three dimensional perspective view of the geographic region comprising: an elevational data storage unit; an imagery data unit; an elevational data processor, said processor including: means for producing an elevational map of said region, said map including an elevational value for each defined coordinate location in a coordinate system; means for creating a set of curves of said elevational values, each curve plotting said elevational value along a single axis of said coordinate system; means for identifying points along each of said curves which have a slope that differ from a neighboring point by a value exceeding a predetermined threshold; means for producing a new elevational map containing only said means for identified points; means for forming tiles in said new elevational map representing elevation features in said region by connecting each said identified point to adjacent identified points in adjacent curves which are less than a second predetermined threshold distance away from each said identified points; and means for generating a three dimensional perspective image of said region from said imagery information produced by said imagery data unit and said elevational map containing said tiles.

11. The system of Claim 10 wherein said coordinate system is a cartesian coordinate system.

12. The system of Claim 10 wherein said means for producing an elevational map of said region further comprises: means for generating multiple visible images of said region from different viewing angles; means for measuring the amount of the apparent shift an features at said coordinate locations between said different views; and means for determining the elevation at said coordinate locations based on the apparent shift.

13. The system of Claim 10 wherein said means for forming tiles forms tiles defined by at least three identified points, wherein lines connecting points form a polygon, whereby said tiles represent elevational features.

14. The system of Claim 13 further comprising means for forming lines between neighboring identified points which do not form tiles, whereby said lines represent elevational features.