EP0526659A1 - Improved map production method - Google Patents

Abstract

Mapping method according to which the data is collected by a remote sensing method such as satellite image data (41) then processed (42) to obtain the representation of a given region. The data corresponding to the geographic characteristics of the region are digitized (43), compared (44) with the satellite image data and corrected for distortion. The satellite image data and the digitized characteristic data are then integrated (48), so that each image data element for which there is a corresponding characteristic data element is modified so as to produce a new combined data element which , after the data integration step, incorporate information relating to the characteristics defined by the digitized characteristic data. The following integration step (42) can provide for the incorporation of text data into the characteristic data and the already integrated image data. The text data can be introduced by placing the text information (49a) on a positioning substrate derived either from image data or contour characteristic data (46) then by digitizing it (51a) or by directly entering the text under electronic form (49b) with a step of processing the associated text information (51b). The final product of the system is in the form of sequential plans (53) suitable for color printing. The specific improvements brought by the process relate to the optimization of image data, which provides a high-quality visual appearance on the map produced, and a distinction between data-characteristics and data-images by a superfine boundary area carried out at the stage of separation preparations for the printing process.

Description

DESCRIPTION

TITLE OF THE INVENTION

"Improved map production method"

r Technical Fieldj

This invention relates to image processing and map production systems. The invention is especially directed to the processing of images or data derived by remote-sensing systems, which may be, for example, satellite-borne or airborne, and in which the images or data are suitably recorded in digital form. In the context of the present disclosure therefore, the terms "satellite image" and "satellite data" are to be understood as encompassing also images or data furnished by other remote-sensing systems including airborne remote-sensing systems, while the term "remote-sensing" is to be interpreted as embracing satellite-borne and airborne imaging and data collection techniques for recording information relating to geographical areas or regions.

•j In particular, the invention relates to the processing

^* * of satellite images, the processing of cartographic data, and to procedures for combining image data and cartographic data with optimum effect, to appeal to the human eye. In the context of the present disclosure, cartographic data are taken to include feature data which relate to information indicating roads, railroads, rivers, crossings and other topographical data and information, usually identified as symbols as in a conventional map, as well as text data. The invention also extends to the processing of cartographic and image data by particular computer graphic techniques and to the integration of data from conventional maps with satellite image data to provide a satellite image map especially suited to multiple reproduction by the relatively economical four colour process printing system.

rBackground Artj

Conventional maps, especially road maps, railway maps, sightseeing guide maps and others, are produced by printing representations of geographical and natural features on materials such as paper along with desired map symbols determined at the discretion of the mapmaker . The geographical and natural features may be expressed by a so- called contour system, a stepwise colouration system, a tonal gradation system, a hachure system, or the like.

Conventional maps are typically based on topographical maps supplied by official surveys, National Geographical

Societies, or other bodies, or they may be based on aerial photographs or satellite images at least for the purpose of indicating geographical and natural features in the maps. Map symbols include indications for ward boundaries, for government and municipal offices, and for schools and the like, as well indications for national roads and the like and identification of railways and the like. It is necessary to establish regulations with respect to the size of map symbols and the thickness of the lines used for them for the purpose of standardizing the indications not only in one and the same map, but also in all maps of the same kind. Regarding the thickness of the lines of the map symbols in question, the finest line thickness employed is about 0.05 mm, but widths of 0.1 to 0.2 mm are also used as representative minima.

Almost all commercial maps based on topographical maps are of the above mentioned type. However, certain kinds of map have more recently been commercialized, prepared more directly on the basis of aerial photographs, or with geographical and natural features derived from satellite imagery. These are hereinafter referred to as "satellite imagery maps".

Where a map is printed on the basis of a corresponding satellite imagery map, symbols and text may be distinguished from the background satellite imagery by adjusting the hue and density of the colours of the symbols or text, so that symbols and text may be distinguished from satellite imagery in the final printed map provided by integration of satellite imagery and cartographic data. However, where only four process colours are used in forming satellite image maps, even making the best use of colour tone and /15080

- 4 -

resolving power, it can frequently be difficult to distinguish the cartographic data, especially road symbols and text, from the background imagery, if the density and hue of the colours of the symbols are merely simply applied. For regions having numerous map symbols of diverse kinds, a means of correcting the background imagery must be employed in preparing conventional maps by using satellite imagery, if clarity of interpretation is to be preserved for the user.

There is known a technique of applying a dropped-out marginal area to the outlines of map symbols, especially letters, for the purpose of distinguishing them from the background imagery. In the case where the outlined symbols are so-called ornate letters of a particular kind, the white and thick outlines around the symbols serve to cover and mask the satellite image and other map symbols. As a result they take away from the special quality of satellite image representation of geographical and natural features by reducing the quantity of satellite imagery data, and in addition they are often offensive to the eye. There thus results the problem that both the quantity and quality of the map symbols and text to be included in the map are noticeably limited. As a conventional means of forming these white and thick outlines, there is a known method of creating a masking plate by forming a wide marginal zone outside the map symbols. A thick glass sheet or the like may be put over a negative film having a pattern print on it, e.g. representing symbol or text data, and a light is diffused through the glass sheet and the negative film, whereupon the outlines of the pattern are thickened, i.e. a marginal area is formed outside the pattern to define a boundary region, in accordance with the pattern on the negative film. This is then used to give a positive, which serves as a mask for forming the white highlight parts. However, with this method there is the inevitable problem that the white marginal zones to be formed are thick and the edges of the outlines formed are not sharp, quite apart from the problems already mentioned above.

Thus, in established and conventional image-making techniques used in order to facilitate users in identifying features in reproductions of satellite images, print and text and other representational material is printed as an overlay on the reproduction of the satellite image.

According to U.S. Patent Specification No. 4,984,279, there is provided a method of making a map comprising the steps of:

(a) processing data collected by remote-sensing to provide image data usable to define a pictorial representation of a selected region, said image data consisting of a multiplicity of elements of data, and each of said elements of data containing image information, (b) digitising a representation of geographical features within said selected region to provide feature data, said feature data consisting of a multiplicity of elements of data, and at least some of said elements of data containing feature-defining information, and

(c) integrating said image data with said feature data to provide combined data usable to define a pictorial representation of said selected region, said combined data consisting of a multiplicity of elements of data,

wherein said image data and feature data integrating step includes

(i) establishing, for each element of image data, whether any corresponding element or elements of feature data exist,

(ii) for any element of said image data to which no element of feature data corresponds, said any element providing an element of said combined data containing only image information,

(iii) for each element of image data for which one or more corresponding elements of feature data exist, establishing for each of said one or more corresponding elements of feature data whether said each of said one or more corresponding elements of feature data contains feature-defining information,

(iv) where said each of said one or more corresponding elements of feature data contains feature-defining information, said each of said one or more corresponding elements of feature data which contains feature-defining information providing an element of said combined data incorporating information pertaining to a feature or features defined by said each of said one or more corresponding elements of feature data which contains feature-defining information, and

(v) where none of said one or more corresponding elements of feature data contains feature-defining information, said each element of image data providig an element of said combined data containing only image information

Suitably each element of said combined data within a defined spacing from another element of said combined data which incorporates information defining an edge region of a feature in a pictorial representation of the combined data is modified to incorporate information defining, in said pictorial representation of the combined data, a surrounding region separating said feature from other information contained in said pictorial representation.

The method of the invention of this U.S. Patent Specification may comprise the further steps of

(1) preparing text for disposition at appropriate locations within said selected region in a pictorial representation thereof,

(2) digitising said text to provide text data, said text data consisting of a multiplicity of elements of data, and at least some of said elements of data containing text- defining information, and

(3) integrating said text data with said combined data to provide printing data including image data, feature data containing feature-defining information and text data containing text-defining information, and consisting of a multiplicity of elements of data.

Said step of integrating said text data with said combined data suitably includes

(i) establishing, for each element of combined data, whether any corresponding element or elements of text data exist,

(ii) for any element of said combined data to which no element of text data corresponds, said any element providing an element of printing data containing only image and/or feature-defining information, (iii) for each element of combined data for which one or more corresponding elements of text data exist, establishing for each of said one or more corresponding elements of text data whether said each of said one or more corresponding elements of text data contains text-defining information,

(iv) where said each of said one or more corresponding elements of text data contains text-defining information, said each of said one or more corresponding elements of text data which contains text-defining information providing an element of said printing data incorporating information pertaining to text defined by said each of said one or more corresponding elements of text data which contains text-defining information, and

(v) where none of said one or more corresponding elements of text data contains text-defining information, said each element of combined data providing an element of said printing data containing only image and/or feature-defining information.

Preferably, each element of said printing data within a defined spacing from another element of said printing data which incorporates information defining an edge region of text in a pictorial representation of the printing data is modified to incorporate information defining, in said pictorial representation of the printing data, a surrounding region separating said text from other information contained in said pictorial representation.

Disclosure of Invention]

It is accordingly an object of the invention to provide an improved method of integrating satellite images with representational data, including text information and cartography, in which image enhancement techniques are used to enhance in optimal manner, the visual result, to match the image data to the cartography and to achieve the best possible visual appeal to the human eye of the integration of satellite imagery and cartography. It is a further object of the invention to facilitate processing of digital image data to match the cartography in various forms which is also processed separately from the satellite image to achieve the optimum relation of imagery and cartography for the most effective visual recognition by the human eye by selecting the optimum hue and density in the final artwork printed by the offset print process.

It is also an objective of the invention to provide an improved method of processing cartographic information (feature data) into a form having the most effective visual appeal in relation to the pictorial representation of the satellite image. The invention thus also has as an object a method of integrating satellite imagery and processed cartography in a four colour offset printing process.

Yet another object of the invention is to provide a map in which map symbols are made easily distinguishable from background satellite imagery under the limited condition of using only four process colours as printing colours.

A still further objective of the invention is the provision of a method or methods affording certain specific improvements in and pursuant to the methods of US 4,984,279 previously identified above.

According to the invention, there is provided a method of making a map comprising the steps of:

(a) processing data collected by remote-sensing to provide image data usable to define a pictorial representation of a selected region,

(b) digitising a representation of selected features within said selected region, with or without text data adverting to said selected features, to provide cartographic data, said cartographic data comprising at least feature data and optionally also text data, and

(c) integrating said image data with said cartographic data to provide combined data usable to produce a printed pictorial representation of said selected region containing image information and cartographic information, wherein said processing step comprises normalising and/or standardising in accordance with predetermined characteristics, information comprised in the image data which defined topographical features in said pictorial representation, to achieve consistency of representation of said topographical features within said pictorial representation for all similar topographical features and/or to achieve consistency of representation of said topographical features between individual maps of a series of maps.

Preferably, said processing step comprises application of at least one of the following operations as defined in the present specification to the image data:

1. Stretch enhancement,

2. Geometric correction,

3. Interpolation,

4. Classification,

5. Contrast stretching, and

6. Edge enhancement, for the purpose of substantially maximising the visual appeal of a pictorial representation of topographical information comprised in the image data.

In a favoured exemplification of the method of the invention, said cartographic data may consist of a multiplicity of elements of data, at least some of said multiplicity of elements of data containing information defining cartographic details, and the method may comprise the following further procedures which follow said digitising step and precede said data integrating step:

(i) processing said cartographic data to provide masking data, said masking data comprising all of the elements of cartographic data which define cartographic details together with a plurality of further elements for defining a region to separate each group of said elements of cartographic data defining a particular cartographic detail from other information contained in a pictorial representation of the image and cartographic data, and in particular, separating said each said group from image data, and

(ii) modifying the image data to provide masked image information in which regions corresponding to masked cartographic details contain no image information.

Said modifying step then suitably comprises the preparation of a plurality of colour separations, each of which is prepared from a corresponding colour separation for the image data and from said masking data, while said data integrating step may comprise bringing together said modified image data and said cartographic data to provide a plurality of colour separations usable in a printing process to produce said printed pictorial representation. It is an especially favoured feature of the invention that each said data element separating an element of cartographic data defining a cartographic detail from other details of said pictorial representation should have a dimension transverse to the boundary of said detail of between 0.0125 mm and 0.075 mm in the pictorial representation of the image and cartographic data, so that the boundary region separating said cartographic detail from other information of the map is substantially undetectable by the naked eye in a printed reproduction of said pictorial representation.

Thus the invention also provides a map comprising an integration of satellite image data and cartographic data, wherein cartographic information is separated at least from image information in a printed pictorial representation of the data by a marginal zone containing no information, the transverse dimension of said marginal zone being between 0.0125 mm and 0.075 mm, so the said marginal zone is substantially undetectable by the naked eye in said printed pictorial representation.

In another exemplification of the method of the invention, said digitising step comprises displaying said image data to provide a pictorial representation of said selected region and deriving said feature data directly from said displayed image by means of a combination of scribing and computer graphics. The technique may alternatively however comprise use of conventional maps to trace the feature data by conventional scribing techniques, storing this information, and subsequently processing it using computer graphic capabilities to produce particular colours and shapes by raster graphics.

In a favoured embodiment of the method of the invention, each element of said combined data either incorporating information defining an edge region of a feature in a pictorial representation of the combined data, or within a defined spacing from an element of said combined data incorporating information defining an edge region of a feature in said pictorial representation of the combined data, may therefore be modified to incorporate information defining, in said pictorial representation of the combined data, a surrounding region separating said feature from other information contained in said each element of said combined data, said surrounding region suitably being provided at the stage of integrating imagery offset films and cartography offset films.

After completing integration of offset films of imagery with those of cartography, including text data, each element of printing data either containing information defining an edge region of text in a pictorial representation of the printing data, or within a defined spacing from an element of said printing data incorporating information defining an edge region of text in said pictorial representation of the printing data, is preferably modified so that it then incorporates information defining, in said pictorial representation of the printing data, a surrounding region separating said text from other information contained in said each element of said printing data. Said text may be laid down on a substrate preparatory to digitisation, or said text may be digitised directly as character data with associated locational and attribute information.

Thus in order to attain the above-mentioned objects, there is provided in accordance with the invention a map formed by combining satellite imagery and cartographic data by four colour process printing which is characterised by white marginal zones around map symbols indicating roads, railways, administrative divisions and other like data, as well as around text data, each said zone being of superfine width such that it is substantially invisible to the naked eye.

The method of the invention may also comprise the further step of making combined offset films of imagery and those of cartography including text to provide one or more representations of said selected region usable for printing purposes.

The invention also extends to a map when made by the method as defined above. In the resulting map, which is formed by printing with four process colours, the map symbols, text, and other cartographic data as printed in combination with the satellite imagery, are sharply and distinctly highlighted because of the non-image area or white marginal zone surrounding said symbols, text or other data. Each said zone has a supefine width such that it is substantially invisible to the naked eye. Said width may be from 0.0125 to 0.075 mm, as applied to the outline of the map symbols. The resulting map is therefore free from the problem of the cartographic data being difficult to distinguish from the background satellite imagery, and, in addition, it has further advantageous merit in that the areas of the background imagery to be masked by the marginal white zones may be reduced to a minimum.

r Brief Description of Drawings]

The method of the invention will now be described having regard to the accompanying drawings, in which:

Figure 1 shows a region for which a satellite picture is to be prepared from portions of four separate partially overlapping satellite images, as indicated in the drawing,

Figure 2 is a detail of an area of Figure 1 where three satellie images overlap, Figure 3 is a diagrammatic representation of the preparation of a traced overlay delineating certain representational features from a printed satellite image,

Figure 4 is a diagrammatic representation of the preparation of text data for association with the satellite image and representational material of Figure 3,

Figure 5 is a flow diagram showing various stages in the processing of data to provide an integrated reproduction of the satellite image for a specific area and associated representational data,

Figure 6 is an enlarged plan view of one embodiment of a map according to the invention,

Figure 7 is a flow diagram showing steps involved in preparing a map according to the invention as shown in Figure 6,

Figures 8 to 11 show steps involved in forming a mask to be used in the preparation of a map according to the invention, and

Figures 12 to 15 are decomposed positives for the map of Fig. 6. In Figures 6 to 15 of the drawings, the reference numerals identify the following features: 101 to 104 are roads, 105 is a railway, 106 is an administrative division, 107 is a string of text, and 110 is a non-printed white part.

r Best Mode for Carrying out the Invention]

The following definitions apply to the descriptive text:

Image data are defined as digital data recorded by satellite, by scanning the earth's surface in a number of spectral bands. Each band corresponds to a different wavelength of reflected radiation recorded by the satellite sensors.

Image data processing refers to a variety of methods of computer processing for enhancement of image data to provide the required final effect by operating on raw satellite imagery.

Feature data refers to those elements of a map which define cartographic features, such as roads, railways, and other symbols.

Text data refers to those elemets of a map which define textual information, for example, indicative of locations or other aspects of either image data or feature data. Certain aspects of the procedures now described with reference to Figures 1 to 5 have already been disclosed in said US 4,984,279. As shown in Figure 1, the area 1 bounded by the double line 2 represents a region for which it is desired to prepare a satellite map by the method of the invention. In order to provide such a map, largely cloud-free satellite pictures or image data must be available for the entirety of the region 1 in question. By reference to indexes of world coverage provided by the satellite companies, it can be established whether or not substantially cloud-free image data is available for the complete area of the region in question. A computer link to the satellite company's data bank facilitates direct interrogation of the satellite company's data to establish whether or not this is the case.

Four notional available image areas 3, 4, 5 and 6, are indicated by the other solid, 3a, dotted, 4a, 5a, and chain- dotted, 6a, areas on the representation of Figure 1. It is assumed that generally cloud-free images are available for each of these. If no more or less cloud-free image is available, then the satellite system may be instructed to record such an image on its next available traverse of the region in question. Alternatively, and depending on the extent of the cloud cover, the cloud data may be processed and merged with the clear-image data. For each of the imaged regions, the data is then delivered on computer tape. For one particular satellite company, three tapes are furnished, each tape providing image data in a particular colour band. These bands may be green, red and infrared. The Landsat system provides up to seven bands, of which blue, green, red and infrared are of interest to satellite map makers. For images or data provided by airborne sensing, different bands may be used.

The computer tapes are loaded and processed on a computer system by the satellite map maker. Each imaged region 3, 4, 5 or 6, may then be displayed on the computer screen, and colours are assigned to the bands, typically in false colour during the initial stages of processing. A multiplicity of known programming methods are available for carrying out these stages of the operation.

In order to build up an image for the required area, as indicated by the double bordered region 1 in Figure 1, data must be used from portions of the various overlapping satellite images 3, 4, 5, and 6, available. In the overlap region 7 of Figure 7, shown enlarged in Figure 2, there is a choice of image data. The various images 3, 4, 5 and 6 may also have different pixel sizes to give different resolutions in each image area. In region 7 therefore, the choice of image source and level of resolution to describe an image portion 18 is left to a skilled operator, using appropriate programming tools on the computer system, to select as between a pixel 13 from image area 3, a pixel 14 defining a region or element of image area 4, or a pixel 16 from image area 6. Similar decisions are applied to other sectors of overlap.

These stages are indicated in the flow diagram of Figure 5. The assembly or collection of satellite image data is indicated by reference 41. This data is then transferred to the step indicated by reference 42, where the satellite data is processed to provide an image of the selected area, for example, in the manner described above in regard to Figures 1 and 2. If a true colour image is required, the next stage is to prepare colour scales equating the various false colours to the required true colours. Again, computer techniques are used, but human intervention and judgement is needed, especially in building up appropriate colour scales and interpreting the information available on the satellite image data. This processing step is not designated separately in Figure 5, but takes place within the steps encompassed by reference 42. If no further steps are involved and the satellite image is then to be printed for use, for example, for poster purpose or the like, then the data is next sent for laser printing onto a negative at this stage. Printing takes place using clour separations derived from the image data to provide the finished product by high quality printing processes. This form of output is not indicated independently in the flow diagram of Figure 5, and takes place within the general ambit of reference 42. Printing at this stage may also be carried out in false colour, and the stages required in adapting the data to provide true colour output may be omitted.

According to specific improvements effected by the present invention, the image data may be subjected to one or more of the following processes:

1. Enhancement stretch

This is a filtering method applied to satellite data in digital form which has the effect of sharpening edges (i.e. boundaries) in the data. Enhancement stretch refers to a stretching technique whereby for any given image, the range of colours measured in terms of intensity of red, green and blue for every pixel in that image representing a particular type of ground cover, is predefined by a set of statistics.These statistics are the result of extensive tests on a given satellite image.

2. Geometric correction

This technique is used to warp a set of digital data ( e . g . a satel lite image ) so that it fits a particular map pr o j ecti on . Gr ound cont r o l points f o r whi ch gr ound coordinates are known from available mapping are identified by pixel in the image . When a sufficient number of control points ar e obtained , a mathematical polynomial can be defined which will fit the image data to the map projection.

3. Interpolation

This is a method of changing the pixel size of a set of digital data. A satellite image is resampled at the new pixel size by calculating intensities of each new pixel based on the intensities of the surrounding original pixels.In practice geometric correction and interpolation can be combined into one step.

4. Classification

This is a method of subdividing a satellite image into a number of classes, each class having a similar range of intensities for each band of raw data. Each class therefore represents a different type of ground cover. Normally there will be classes corresponding to water, shadow, forest, grass, crops, rock and urban areas. Classification is fundamental to the method of the invention, as the aim is to ensure that similar types of ground cover have similar spectral characteristics (i.e. colour) in the final image, regardless of location or time of year.

5. Contrast stretching

This is a method of converting a given range of intensities (i.e. those of the raw satellite data) into a new range with the aim of increasing the contrast in the data and/or obtaining a required range or colours. In the specific case of the invention, the required range of intensities are pre-defined for each class of ground cover type, as a result of extensive tests on an original test area image.

Statistical analysis of the final output of test area images enables a mean and standard deviation of the intensities of all pixels within each class to be calculated. Enhancement stretch as hereinbefore described may therefore be defined as that contrast stretch which when applied to a satellite image regardless of the location of the image or the time of year, provides an output image which has the same range of intensities in terms of red, green and blue.

6. Edge enhancement

This is a filtering method which exagerates or enhances natural edges in the data (i.e. the effect is to sharpen all boundaries in intensity values) and so to increase the apparent resolution. The method involves the comparison of each pixel with the average value of the surrounding pixels and an exaggeration of the resulting differences. The size of the filter or kernel is defined as the size of the nearest neighbour matrix surrounding the central pixel. The kernel size may be 3 x 3, i.e. 3 pixels by 3 lines. The net result of application of these various techniques, either in combination or in isolation, is to provide a method of reproducing satellite image data in a highly consistent manner, approximating more closely than has hitherto been attainable, to the manner of representation of information in a map prepared by more traditional means. In particular, the colour treatment and intensity control features available by virtue of the present invention enable more diffuse colouration of large topographical features such as areas of water, i.e. lakes or the sea, to be achieved, thereby providing a significantly enhanced and substantially optimised visual effect. The invention thus enables advances in the use of suitable data and its integration with cartographic information to produce maps of consistent quality and appearance offering convenience of use to the user combined with an attractive visual appearance.

A particular feature of the system of the invention is therefore the integration of traditional map data with satellite image data. Alternative techniques, both of which are indicated on the flow diagram of Figure 5, are available for carrying out this step. In the first such technique, a traditionally prepared map of the region to be imaged is digitised, reference 43, so that all of the information to be printed on the finished satellite map is stored in computer - r eadab 1 e form. Depending on circumstances, this digitised data may or may not be directly usable for integration with the satellite image data. Most usually, at least some correction of either the digitised map data and/or the image data for certain distortions will be required, reference 44, so as to ensure that when the map data is printed onto the satellite image, there is correct registration of the digitised map features with their images as recorded by the satellite system. The techniques adverted to previously relate inter alia to treatment of this kind of the image data. Distortions exist in traditional maps for a number of reasons, including map projection techniques, which lead to deviations particularly at the edge regions of maps, where the curved surface of the earth is projected onto the flat surface of the map page. Distortions also exist due to the exaggerated dimensions of conventional symbols such as roads on traditional maps. Exaggeration is however necessary so that the features in question can be distinguished. Further distortion may exist in the satellite image itself in that the image may not necessarily be recorded from a position vertically over the region photographed. In other words there may be a slight angling effect, which also has to be compensated before the map data and the image data can be combined or integrated. In general, if the quality of the map is good, then it will often be preferable to adjust the image data, but where map reliability is poor, then the image data will take precedence. Accordingly, an alternative system of preparing the data for the satellite map is to initially print out the image data in precisely the same size as the intended final size of the satellite map to be produced, reference 45. This printed image is then compared against a traditional map of the region and the roads and other features requiring to be represented on the finished satellite map are traced off the satellite image onto a suitable transparent material, reference 46. Features such as roads are usually clearly identifiable in high definition satellite images, although certain other features may be less discernible, but features such as county or district boundaries which have no physical reality on the ground but may require to be shown on the map, must be located by reference to identifiable features on the satellite image. It is important that both the satellite image and the tracing prepared from it are on dimensionally stable materials, so that no additional distortions are introduced at this stage. When the tracing 46 is prepared and all of the features requiring to appear on the finished satellite map have been included on it, the traced data is then digitised, reference 47. By proceeding in this way, the digitised features are in a format which is fully compatible with the image data, and difficulties of registration are minimised.

It is also possible to "trace" the map data directly off a screen display, i.e. to cause a cursor to traverse a path following a map feature to be digitised and to record the path followed by the cursor. This may be carried out using appropriate software in conjunction with a tracker ball, mouse or graphics tablet. Alternatively a light pen may be used, working directly on the screen to move along the path defined by the imaged feature. In this way, the step of physically tracing from a print of the image may be avoided, and feature data digitized directly. In this variant, step 45 is bypassed, and steps 46 anf 47 take place substantially simultaneously from cursor or light pen path data.

The foregoing method involving tracing from a printing of the satellite image is preferred for more complex or large-scale maps, but for relatively small-scale satellite maps, direct digitisation of the map and subsequent correction, at stage 44, by computer /manual input is generally acceptable. In regard to the "tracing" method, Figure 3 shows in schematic form, on the right hand side, a satellite image 21, and, on the left hand side, a tracing 22 prepared from it showing a portion of coastline 23, some roads 24, a river 25 and an airport 26.

The digitised data 47 from the tracing 46 is then "burned in" or integrated fully into the satellite image data, reference 48, Figure 5. The features recorded digitally in conventional mapping terms, step 47, are not applied to the satellite image from step 42 as an overlay. A computer controlled process takes place in which the digitised data 42 for each region of the image, at pixel level in display or printing terms, is compared with the digitised map feature data 47, and where this latter contains feature-defining information, it takes precedence, together with an appropriate bounding or boundary region or area, over the original satellite image at 42, to ensure that in the finaly printed satellite map the mapped features 47 will not only appear but will also stand out against the background of the satellite image 42.

The first step in the operation is to establish whether or not any digitised map feature data exists for the particular pixel or element of image data being so compared. If the feature data is also in pixel format, and the respective pixels are of the same dimensions, then a straight pixel by pixel comparison may take place, and where the feature pixel contains feature-defining information, it completely replaces the corresponding pixel of image data to provide an element of the combined or integrated data resulting from this stage of the process of the invention. Any pixels of satellite data within a predetermined distance outward of a feature pixel containing feature-defining information are likewise repalced, but in this instance by a white or non-information containing pixel, to define a boundary zone surrounding the feature information in a display or reproduction of the combined, integrated or "burnt-in" image and feature data. Thus, in these circumstances, the satellite image data 42 may be essentially wiped out or erased underneath the merged or burned-in map features 47 incorporated at this stage, and also over a boundary or surrounding region in the printed pictorial representation of the data, so that the features will stand out in a clearly distinguishable manner from the image data in a finished satellite map prepared on the basis of the combined or integrated data. Where the image and feature data sets are provided by different densities of pixel, then the combined or integrated data is given by first of all bringing the respective resolution of the data sets to a common value. Thus in this circumstance, the image data set must initially be modified so that the final data set following the integration step can incorporate the feature data as required. Hence, if the feature data has smaller pixels than the image data, a number of pixels of combined data will exist for each initial pixel of image data, but only some of these pixels may have been replaced in the integration step, the remainder continuing to provide similar information to that cotained in the original larger pixel. This operation of replacing an element of image data by a plurality of successor elements, some of which may remain unaltered, but others of which may be replaced or superseded by new data, is referred to herein as modification of image data, and modification of image data is to be understood in this sense. This provision of, according to the present invention, a surround or boundary region in digital data by modifying digital image data, or by over-riding or erasing or over-writing the digital data, differs from the defining of such a feature by modifying tint data. Where the feature data elemet does not contain feature-defining information, or where no element of feature data corresponds to the element of the image data in question, then the element of image data remains unchanged and forms an element of the combined or integrated data, or is used to provide or derive a plurality of elements of the combined data, the latter where a transformation of pixel size is in question. Such a transformation may also result in an element of image data contributing to the information content of an element of combined data of larger size or pixel dimensions.

Where the pixel sizes of the image and feature data differ, or where the feature data is defined in terms of attributes (e.g. lines represented as curves or polygons defined in terms of co-ordinates rather than on a pixel basis), then modification of the image data or pixels may again be in question rather than any kind of pixel-by-pixel comparison and replacement, for example, by assessment of the effect or impact of attribute or vector data on the relevant elements of the image data, to provide the combined data achieved by the data integration step of the invention. In both of these situations and also in regard to pixel by pixel comparison or replacement, proprietary commercial programming techniques and programs are available to carry out the steps and operations required, and in particular, in the case where the feature data is in non- pixel form, to determine from the digitised information or data defining the feature data, whether or not any particular element or pixel of image data requires to be modified, or replaced, and, if modification is in question, the nature and extent of the modification necessary.

The next stage is to prepare text data, reference 49, for similar integration with the combined map feature data and satellite image data in a further burning in or data integration stage. Again two methods are available. In the first such method, the text required is typeset and laid down on a film or other substrate at appropriate locations. This may be done manually. The text may be laid down while the substrate is overlaid on the tracing showing the map features, so that the lettering will appear in precisely the right place. Figure 4 shows the items already depicted on Figure 3 together with the further material necessary to prepare text data 31, on the left hand side. The locations 32 of text are indicated, and also shown in dotted outline are the map features 23 to 26 of the central compartment of Figure 4, so that the location of the lettering adjacent to particular features can be clearly seen. In fact, these features, depicted in the left-hand compartment of Figure 4 in ghost outline, do not appear on the text preparation substrate in practice, and are shown here only to explain the principles of the invention. The typeset text 32 is surrounded by appropriate boundary regions , not shown per s e in the drawing , ag ain f or distinguishing between the text and the background.

Following preparation of this text material , it is digitised , reference 51 for amalgamation or burn-in in single pixel or other mode to provide a final assembly of computerised printing data suitable for preparing for preparing separations for printing . Again , the data is integrated, step 52 , with the computer data containing both the map features and the satellite image material in such a way that the text data takes precedence over everything else, whether map feature or satellite image . In this way, any overlay effect is again avoided , and the text stands out in the finished product. Thus the combined feature and image data is modified or wiped out or erased underneath the merged or burned-in text data and also over a boundary or peripheral region surrounding the text , so that the test will stand out in the finished satellite map. Precisely the same techniques of image data pixe l repl acement or modification, as already outlined in regard to the "burn-in" of feature data , also apply to the integration of text data.

These so-cal led "burn-in" stages use commercial ly available software , which is capable of combining material of different resolutions . Thus , as already noted , data replacement in the original image data is not necessarily one-for-one on a pixel basis . The image data may be , for example, 6 pixels/mm, while text or feature data may be, for instance, 38 pixels/mm. The programmes used allow combination of data of different resolution while also providing aesthetically pleasing results.

The alternative method of preparing text data involves assigning co-ordinates for each portion of text data to be inserted on the map. The data is then set up in conventional ASCII form, and additional computer data characters associated with each batch of text, namely to define the attributes of the text such as font, size, framing and so on. This data consisting of ASCII strings and attribute material is then converted into vector strings, which are rastered or scanned to provide digitised text data for burn-in onto the integrated image and map data to provide the final pre-printing computer data 52 defining the satellite map. As in the case of the feature data, the commercial software used enables comparison of the text data with the data from a previous stage of the system or method of the invention, namely image data or combined or integrated image and feature data, and allows modification or replacement of that previous data as required, whether on a pixel by pixel basis, or by establishing from the content and attribute information of each element of text data, the need or otherwise for modification or replacement of an element of the data from the previous stage, and, if modification is required, the nature and extent of such modification. The resulting data may be outputted as a Chromalin (Trade Mark) to provide an advance indication of the aticipated finished product, or a machine proof may be prepared.

The final stage 53 makes use of this integrated computer data to prepare a number of colour separations for use in the printing process. Typically four separations are used, cyan, yellow, magenta and black. Dimensional stability is exceedingly important to ensure a very high quality of output, as is the quality of the paper used for the final print product. The output of the final printing is a so-called satellite map, consisting of a satellite image with, integrated into the satellite image, conventional map features such as roads, airports, public locations, railways, and other communications features, along with text data identifying these features and other geographical locations. The system of the invention ensures that the finished product does not in any way resemble the result of overlay printing, and ensures an extremely high quality of integration and representaton of map data on the satellite image. The system of the invention combines high quality with a large choice of colours, conventions as to width, key lines, casings and feature sizes etc., and, in regard to text, a choice of, inter alia, fonts and size, to an extent not capable of being achieved using overlay printing.

Certain further aspects of the present invention will now be described having regard to Figures 6 through 15, of which Figure 6 shows an embodiment of a map according to the present invention.

Figure 6 is an enlarged view of a general road map, where 101 is an expressway, 102 a national road, 103 a main road, 104 another road, 105 a railway, 106 an administrative division boundary between, for example, a metropolis and outer districts, or the boundary of a city, town or village, 107 text data indicating the names of prefectures, places and roads, 108 a river, and 109 a region of mountains, forest, hills, or town. The regions delineating rivers, mountains, forests and the like are derived from satellite imagery.

Specifically by creating non-printed image area between map symbols and text data (letters) on the one hand, and background satellite imagery on the other hand to form a superfine marginal zone of a width between 0.0125 mm and 0.075 mm, the map symbols such as the expressway 101, the national road 102, the main road 103, the other road 104, the railway 105, the administrative division 106 and the letters of text data 107 may be severed from the background satellite imagery and made to stand out. An example of such a superfine marginal zone consisting of non-printed image area is indicated by reference 110 in Figure 6.

In Figure 6, the edges of the outlines of background satellite imagery and map symbols or text data are shown as line drawings for representational purposes, but in an actual map, these edges are generated by screen dots of colour separation or flat screens. A method of preparing a map according to this invention will now be explained in detail with reference to Figures 6 through 15 of the accompanying drawingsin which integration of satellite imagery and cartographic data (map symbols plus text) is used to provide four colour process films ready for offset printing. The production of screened positive films providing colour separations to be used for printing employs a digital data processing system including a computer (for exmple, the PASTE 4000S system of DAI-NIPPON SCREEN CO.). Referring to the flow diagrams and schematic system layout of Figure 7, the computer system includes a sc'anner 121, a host computer 122, a graphic work station 123, a magnetic tape controller 124 and a laser plotter 125.

The preparation of the map includes the following steps:

1. First, a single-stripe line (0.1 mm) is scribed to indicate the centre of a road or other feature in accordance with the original feature data of the selected region. This is then turned into a positive film to be used as original artwork 120 as a film mechanical. 2. The original artwork (positive film mechanical) is scanned, 121, for conversion into digital data (dot information) .

3. The converted digital data are then automatically formed into core lines (width 0.025 mm = 1/40 mm) at graphic work station 123.

4. The core lines of the data are next classified into type of map symbol, such as expressway 101, national road 102, main road 103 and other road 104 with respect to the data layers and then formed into two-stripe lines. This two- stripe line information is stored as feature data information for final map making. Thereafter the peripheral areas of the resulting two-stripe lines are thickened by applying outlines of one pixel width thereto to make a separate film plate for masking purposes (masking plate).

5. The masking plate containing all the map symbols of roads, railways and other data expressed by two-stripe lines for the respective data layers thickened by one pixel unit, is then photographed onto film and developed, 126.

6. After developing, the resulting positive film of the masking plate is applied to each screened offset negative film for the four process colours - cyan, magenta, yellow and black respectively - corresponding to the four colour separations of satellite imagery, along with non-exposed films. After proper exposure of these latter films to light, they are developed. The resulting negative films are then turned to positives in which the masking plate areas show blank, so as to appear as dropped-out white on finally printed matter.

In preparing the map by the method of the invention, the step of forming the masking plate is important. This step is exclusively conducted in the graphic work station (GWS), item 123 in Figure 7.

This step will be explained in detail with reference to Figures 8 through 11:

Firstly, with respect to the data as formed into core lines (Figure 8), the data are classified into layers for the various kinds of lines, such as expressway lines, national roadway lines, etc.

Next, the core lines are formed into two-stripe lines by any suitable treatment. This is effected for each of the respective data layers on the basis of the characteristics previously determined or prescribed in respect of the width of roads and the thickness of roadside lines (Figure 9).

Referring now to Figure 9, after all the lines B through G have been scribed, i.e. core lines B, D and F and two-stripe or side lines C, E and G, each of the two-stripe or side lines C, E and G is further scribed by one pixel unit, as shown in Figure 10. These scribed outlines are designated H, I and J respectively. This results in a thickening or widening of the two-stripe or side lines defining the symbol in question. In the present example, one pixel is set at 0.025 mm = 1/40 mm. However, it may also be defined to be 0.0125 mm (= 1/80 mm). The lines may also by thickened by an integer number of times of one pixel as appropriate. The resulting masking plate comprising lines B through J will provide the mask for use in preparing the map according to the invention (Figure 11).

When the positive film containting the map symbols of roads, railways, etc., formed as two-stripe lines for the respective data layers, i.e. essentially as shown in Figure 9, is placed over the positive film as derived from the masking plate i.e. essentially the representation of Figure 11, and the two attached positives are observed with the naked eye, only the parts as thickened along the outlines of the two-stripe lines for the respective data layers are seen, projecting outwardly beyond the outlines of the two- stripe lines of the masking plate. Using this principle in the positive films for the respective colour separations of the imagery corresponding to the four process colours, the parts of the image to be replaced by map symbols are made transparent by the use of the masking plate, and these parts which are made to be transparent or white in accordance with the outlines of the two-stripe lines are also thickened by one pixel unit therearound. Accordingly, since in the final production process, the two-stripe lines for the respective data layers are synthesized as two-stripe lines bounded by transparent and white part, non-printed parts or white parts each having a superfine width of one pixel unit or more pixel units as appropriate are formed along the outlines of the printed two-stripe lines.

The reason why the superfine width of the non-printed parts or white parts is defined to fall within the range of 0.0125 to 0.075 mm is as follows:-

When the width is broader than 0.075 mm, the white colour of the two parts would be too clearly distinguished from the other parts and the satellite imagery would be reduced because of such broad outlines. As a result, those characteristics of the satellite imagery which are close to the actual geographical and natural features would be diminished, and the value of the resulting map would therefore be reduced. On the other hand, when the width is less than 0.0125 mm, the superfine non-printed parts or white parts can hardly be formed around the outlines of the two-stripe lines because of printing precision errors.

When the map of Figure 6 is printed by the use of four process clours - cyan, magenta, yellow and black - four printing plates are prepared using a positive offset film for the black plate of Figure 12, a positive offset film for the magenta plate of Figure 13, a positive offset film for the yellow plate of Figure 14 and a positive offset film for the cyan plate of Figure 15. Using the four printing plates thus prepared the intended map is formed by a determined offset printing step.

The map according to the invention, which is formed as mentioned above, is obtained by combination of the three techniques of the present invention involving use of a digital data processing system to freely scribe the desired images and letters at the discretion of the scriber at one pixel unit, the use of image processing techniques to convert the results of remote sensing data to natural colours familiar visible to the human eye, and practical knowhow relating to the preparation of maps.

A map according to the invention is thus prepared by the above-mentioned method. Specifically, map symbols having superfine non-printed or white outlines in a map comprising basically four principal colours are clearly distinguishable from the satellite imagery showing background geographical and natural features. As the outlines have a superfine width which is invisible to the naked eye, the parts of the background satellite imagery masked by the map symbols are reduced to a minimum, so that the accurate and detailed characteristics of the geographical and natural features shown by the satellite imagery may well be maintained in a map according to the invention.

Additionally, descriptions of map symbols which need to be written in larger amounts and with great accuracy may be effected by the method of the invention without significantly interfering with the characteristics of the background satellite imagery. Accordingly a map which is highly practicable, easy to read, and easy to use, can be produced . The present invention therefore provides for a separate handling and processing of image data and cartographic data until a relatively advanced stage of data processing. Integration of processed image data and cartographic data then takes place only at the last stage of the operations to enable preparation of the four-colour separations used in printing the finished map, which contains the processed image and cartographic data.

Satellite image data is digitised in known manner. Cartographic data is prepared as described in the foregoing text and digitised and then adapted to provide masking data by the additon of the masking feature. The masking data not only provides the cartographic information for the final map but also serves to blank out the image data in those regions which are to be occupied in the finished map by cartographic information, together with a narrow region which separates in the finished map the cartographic information from image information or underlying cartographic information, thereby providing an enhanced visual effect. This narrow marginal zone is precisely defined and its dimensions are limited to values suitable for achieving the desired visual appearance while yet minimising the quantity of image information which is obliterated or lost.

Separate handling of the image and cartographic data enables the normalisation and optimisation procedures previously described to be carried out in a particularly effective manner. Thus, for the image data, the colouration in particular of topographical features can be finely tuned and adjusted to provide precisely a desired effect, without this aspect of the processing being impinged upon by the cartographic data. The cartographic data can be likewise optimised in isolation from the image data, in respect of colouration, dimensions, layering and other relevant aspects.

Thus, in one specific procedure in accordance with the invention, the following steps are carried out:

1. The image data is processed to provide representational optimisation, as described above.

2. The image data is used to provide three or four colour separations for use in the printing process.

3. The cartographic data is prepared and optimised, independently of the image data, as described above. 4. The cartographic data is then used to prepare a masking plate.

5. This masking plate is applied to each of the three or four colour separations of image data to provide further separations defining masked image data, in which areas to be occupied by cartographic data and the surrounding marginal zones, in the final printed map, are free of feature- defining information.

6. A plurality of colour separations are prepared from the cartographic data.

7. The masked image separations and the cartographic data separations then define combined printing data within the meaning of the invention and are integrated or brought together in the printing process to provide the final printed map.

r Industrial Applicapabilityj

As described in the foregoing text, the masking or obliteration of those portions of the image data to be occupied by cartographic data in the finished map involves the preparation of a masking plate, involving a four-colour separation process to define this maskig plate. The masking plate is applied to the image data to achieve the required effect. However, it will be appreciated that a similar result could be achieved by wholly digital or alternative means, as appropraite.

Claims

1. A method of making a map comprisig the steps of:

(a) processing data collected by remote-sensing to provide image data usable to define a pictorial representation of a selected region,

(b) digitising a representation of selected features within said selected region, with or without text data adverting to said selected features, to provide cartographic data, said cartographic data comprising at least feature data and optionally also text data, and

(c) integrating said image data with said cartographic data to provide combined data usable to produce a printed pictorial representation of said selected region containing image information and cartographic information, wherein said processing step comprises normalising and/or standardising in accordance with predetermined characteristics, information comprised in the image data which defines topographical features in said pictorial representation, to achieve consistency of representation of said topographical features within said pictorial representation for all similar topographical features and/or to achieve consistency of representation of said topographical features between individual maps of a series of maps. 2. A method according to Claim 1, wherein said processing step comprises application of at least one of the following operations as defined in the present specification to the image data:

1. Stretch enhancement,

2. Geometric correction,

3. Interpolation,

4. Classification,

5. Contrast stretching, and

6. Edge enhancement,

for the purpose of substantially maximising the visual appeal of a pictorial representation of topographical information comprised in the image data.

3. A method according to Claim 1 or Claim 2, wherein said cartographic data consists of a multiplicity of elements of data, at least some of said multiplicity of elements of data containing information defining cartographic details, and the method comprises the following further procedures which follow said digitising step and precede said data integrating step: (i) processing said cartographic data to provide masking data, said masking data comprising all of the elements of cartographic data which define cartographic details together with a plurality of further elements for defining a region to separate each group of said elements of cartographic data defining a particular cartographic detail from other information contained in a pictorial representation of the image and cartographic data, and in particular, separating said each said group from image data, and

(ii) modifying the image data to provide masked image information in which regions corresponding to masked cartographic details contain no image information.

4. A method according to Claim 3, wherein said modifying step comprises the preparation of a plurality of colour separations, each of which is prepared from a corresponding colour separation for the image data and from said masking data.

5. A method according to Claim 4, wherein said data integrating step comprises bringing together said modified image data and said cartographic data to provide a plurality of colour separations usable in a printing process to produce said printed pictorial representation.

6. A method according to any of Claims 3 to 5 , wherein each said data element separating an element of cartographic data defining a cartographic detail from other details of said pictorial representation has a dimension transverse to the boundary of said detail of between 0.0125 mm and 0.075 mm in the pictorial representation of the image and cartographic data, so that the boundary region separating said cartographic detail from other information of the map is substantially undetectable by the naked eye in a printed reproduction of said pictorial representation.

7. A map comprising an integration of satellite image data and cartographic data, wherein cartographic information is separated at least from image information in a printed pictorial representation the data by a marginal zone containing no information, the transverse dimension of said marginal zone being between 0.0125 mm and 0.075 mm, so the said marginal zone is substantially undetectable by the naked eye in said printed pictorial representation.

8. A method of making a map substantially as described herein with reference to any of Claims 1 to 6.

9. A map prepared by the method of any of Claims 1 to 6 or Claim 8.

10. A map substantially as described herein with

reference to Figure 6 of the accompanying drawings.