Map With Indexes for a Geographical Information System and System for Applying Same

The present invention generally relates to geographical information systems and, more particularly, to maps of the type bearing a graphic representation of cartographic information, broken down into a large number of elementary areas with which are associated respectively a large number of indexes.

At the present time, two different approaches are used in cartography, between which there is generally no interpenetration.

The older approach, consisting in the use of maps bearing a graphic representation, is still very widely used, both in professional applications and in popular applications. The maps give global information that is directly accessible and familiar to all potential users. They remain easy to handle even when produced in large formats, in the order of one square meter. When they are made using modern printing techniques, which have now been thoroughly mastered, very fine lines can be produced on them: many maps available at low prices are produced using polychromy by superimposing frames, each formed by a matrix of dots with a pitch of 100 .mu.m or less (down to 25 .mu.m), and with a dot positioning accuracy of approximately 10 .mu.m.

Counterbalancing these advantages, the conventional graphic representation type map has certain limitations. In particular, the amount of useful information that can be inscribed on a map without detriment to its legibility is limited. This often leads to the greater part of the information being placed in an appendix. The text and graphic information are generally interlinked by means of a simple squaring system, whereof each square is identified, by means of a letter-figure pair or by a number. Such squaring gives only an approximate indication and, in addition, obliges the user to refer to indications provided in the margins of the map (see document FR-A-2 639 452).

The data processing approach to geographical representation has essentially consisted, hitherto, in defining an image, dot by dot, in the form of a table, in a memory. Even if use is made of representational algorithms in a simplified form, the necessary information takes up very large volumes of memory, which can amount to as much as several hundreds of millions of bytes for a single map. In addition, existing visual display devices are very far from offering the same degree of resolution as the conventional graphic maps. A professional quality high resolution monitor displays an image in the form of pixels each of which can be likened to a square with sides of approximately 300 .mu.m.

The present invention aims to provide a map of the type defined above that can be used in a geographical information system to combine the graphic qualities of the conventional map with the advantages conferred by the electronic addressing of a data base, this being achieved without detriment to the legibility of the map.

For this purpose, each index comprises a group of dots placed on the map in the vicinity of the associated elementary area and forming a specific code for the said area, the indexes as a whole forming a location frame covering the graphic representation and superimposed thereupon, each group of dots being readable by optical reading means cooperating with a data processing system comprising a data medium that is addressable by means of the indexes.

Each location index serves as a key giving access to information that is not provided on the map, and which may differ according to the intended use of the map and can easily be updated. The entire text of an appendix, all the logical and topological relationships, as well as all the calculation formulae, can be held on a data processing data medium belonging to a calculator of conventional architecture, possibly pocket-sized. The memory volume is, indeed, very small as the calculator does not have to store any of the graphics functions or the most voluminous data.

For reading purposes, the data processing medium can be addressed very simply using optical reading means which it suffices to place on the map at the location where the information is sought for. Such reading means can simply take the form of an optical wand reader having a scanning field representing 2 to 5 times, in each direction, the overall dimensions of an index. Although various encoding modes are possible, it seems to be of particular interest to use an index constituted by a rectangular block of dots printed along several rows of the same length. These rows are advantageously separated by empty rows, and it is also advantageous to encode only one out of every two pixels of the frame in a row, the intermediate pixel remaining empty. Furthermore, one out of every two blocks is left empty, the blocks of coding then forming a discontinuous frame resembling a checkerboard which facilitates the identification and decoding of the block the best centered in the scanning field of the wand reader, should several blocks appear therein.

Numerous security inks are already known which have a response, that is to say a brilliance, which is very weak when they are illuminated by visible light, and which, on the other hand, have a strong response (reflectivity or fluorescence) in a given line of the spectrum when they are illuminated by a light outside the visible spectrum, generally in the ultraviolet range. The wand reader then comprises means for locally illuminating the map by means of a light that excites the security ink and a filter designed to isolate the response wavelength of the ink.

The invention will be more readily understood in the light of the description that follows of particular forms of embodiment, given by way of non-limitative examples, and with reference to the accompanying drawings, wherein:

FIG. 1 is a representation, on a very large scale, to bring out the frame structure, of a fragment of a conventional geographical map along a coastline (the colours cyan and yellow being represented respectively by dots containing crosses by and by empty dots).

FIG. 2 schematically shows a possible representation of code 16384 with the help of a block of 10.times.10 pixels, on a considerably enlarged scale.

FIG. 3 shows the representation of several figures in the "three out of five" code used for the representation in FIG. 2.

FIG. 4 shows, on an enlarged scale, a region in which the location frame has a succession of indexes in a checkerboard pattern.

FIG. 5 represents, without any enlargement, a portion of a map on which is superimposed the location frame according to the invention.

FIG. 6 is a variant of FIG. 2.

Conventional geographical maps are composed of matrices of dots distributed according to a regular frame. In the frequently occurring case of maps produced by tetrachromy, three colour frames are used, cyan (for the areas of water), green (for the vegetation) and yellow (for the desert areas), by superimposing which additional shades can be obtained, and a black frame, in particular for the inscriptions.

In the conventional printing process, each frame takes the material form of a mask on a photographic film the transparency of which is variable according to the dots, or offset film. FIG. 1 shows, by way of example, a map fragment, considerably enlarged to bring out the dots in the frame. On one side of the coastline, in a desert area, the dots are yellow while, on the other side, they are cyan.

There are many other methods of manufacturing maps, some of these better adapted to special applications. It is possible, in particular, to print maps using an ink-jet printer or an electrostatic printer, particularly when the indications are frequently modified in order to adapt them to a particular assignment. The invention is also applicable to the case of a graphic representation that was originally continuous, but subsequently framed (an aerial photograph, for example). In the case of digital photographs obtained, for example, from airborne scanners or observation satellites and reproduced using photo-restorers, the invention is particularly applicable thanks to insertion of the location frame directly in the digital files, prior to restoration.

To implement the invention, disjoined codes formed on a frame that is superimposed on the graphic indications are superimposed on the graphic representation dots. Conventional printing processes make it possible, without any difficulty, to print dots of 50 to 100 .mu.m in diameter located to within better than 10 .mu.m inside a pattern. Use can be made, in particular, of a square pattern of approximately 1 millimeter.times.1 millimeter formed by dots of approximately 100 .mu.m in diameter. To facilitate identification and decoding, it is advantageous to adopt a pattern formed by rows of encoding dots separated by empty rows, each location intended for an encoding dot being separated from the adjacent locations by an empty space.

In the case contemplated above, involving a 1 millimeter.times.1 millimeter pattern, this means that there are five rows available, each having five encoding locations and five spaces. Use can be made, in particular, of a binary code for identifying the FIGS. from 0 to 9 of the "3 out of 5" type, indicated in FIG. 3, including three encoding locations 30 on which is placed a dot, and two encoding locations 31 without dots. The figures thus encoded are indicated in FIG. 3, to the left of each row.

The expression "3 out of 5" covers the opposite case in which the three locations 30 are left empty and the two locations 31 contain a dot. This code has the advantage of facilitating the location of an index, as each row comprises the same number of dots having the same radiometric value, that is to say the same light intensity in the colour in which they are represented. FIG. 2 shows such an index, 10, wherein, for each encoding row, the two empty encoding locations 31 have the same brilliance as the dots in the frame not belonging to an index and relating to the graphic representation frame of the map, whereas the encoding locations 30 containing the encoding dots are of greater brilliance. In the example given in FIG. 2, the index 16 384 is represented on a pattern with five "active" rows and five locations per row. Five rows per pattern suffice to encode 100,000 positions.

The index in FIG. 2 is thus constructed on the basis of an elementary matrix of pixels composing a matrix of larger dimensions, which forms the photosensitive face of an optical wand reader, the dots disposed on the elementary matrix forming a binary code of certain active pixels in this matrix.

It will be noted that the square index in FIG. 2 is substantially isotropic, dots 30 being distributed approximately in all directions about a centre of the index, which is also apparent from observing the different indexes 10 in FIG. 4; this arrangement permits quasi punctual encoding of a map. Conversely, an index having a form as anisotropic as a line of dots, would be less appropriate for the desired encoding. More generally, a matrix comprising a similar number of lines and columns will be considered as substantially isotropic.

Another example of an isotropic index is represented in FIG. 6, where the active "rows" are, here, concentric circles 11 to 13, separated from one another by inactive circles, dots 30 and blanks 31 corresponding to those in FIG. 2 being identified.

It will be noted that the density of the dots on the index in FIG. 2 is low, as it comprises fifteen dots for one hundred pixels. This choice contributes to making the location frame on the map difficult to see with the naked eye. More generally, "low density" should be taken as meaning a density of less than 20 dots for one hundred pixels in the case of an encoding frame that is visible to the naked eye, and of less than 50 dots for one hundred pixels in the case of an encoding frame that is invisible to the naked eye.

If a security ink, not visible to the user, is employed, the appearance of the map is not impaired; the indexes are easy to isolate using a reader; each index can be superimposed directly on the dot to which it relates.

In those areas of the map in which the different indexes are very numerous, each index can be printed once only exactly at the point concerned. In most cases, the blocks encode the pair of cartesian coordinates of their location; thus all the blocks in the location frame indicate a different code.

To facilitate further the identification and decoding of the pattern best centered in the scanning field of the optical reading means, it is possible to incorporate in the frame only one block out of two and to arrange the incorporated blocks after the fashion of a regular rectangular checkerboard, as shown in FIG. 4. The blocks can further be indexed by adopting, as a block indexing scheme, a curve having the property of filling the entire square and thus of permitting the definition of a dot with a single parameter. Among such curves, mention can be made, in particular, of Hilbert's curve, a description of which is to be found in the article entitled "Fractals et dynamique des iterations" (Fractals and the dynamics of iterations) by Claude BREZINSKI, AFCET/INTERFACES No. 88, February 1990, page 3.

It will be noted that the arrangement of the dots of the index in FIG. 2 in rows and columns facilitates identification, as these define two orthogonal directions, x,y (FIG. 4), enabling the map to be orientated about the index in question. What is more, the arrangement of the indexes in relation to one another can also provide this identification: the checkerboard arrangement of indexes 10 in FIG. 4 again defines the two orthogonal directions, x, y.

The data processing means associated with the map comprise a mass memory containing a data base and addressable by means of the index, a processing unit enabling the data base to be consulted using an input means such as a keyboard, and an alphanumeric display means. Addressing to obtain information corresponding to a given index is carried out by means of opto-electric means such as a wand reader with a matrix of charge coupled sensors having a field the diameter of which generally corresponds to about three times the dimension of the pattern. When the patterns are of the special constitution described above, the wand reader can have a field of 5 millimeters in diameter and comprise focussing optics and a CCD camera. The processing unit is programmed so as to carry out simple morphological processing for individualizing and then decoding the index. The processing unit can be supplemented by an interface linking up with a global positioning system receiver, now available on the market, and enabling the position of the receiver to be determined to within ten or so meters, with reference to the position of twenty-four satellites in orbit.

The processing unit can be designed in order to permit the addressing of data in the memory by means of keywords or standard questions, so as to make it possible to read the answers to questions, even complex ones, such as:

the number of hotels in the town indicated by the index at which the wand reader is pointed,

towns with over five thousand inhabitants at less than one hundred kilometers by road in the administrative region at the police headquarters of which the wand reader is pointed,

direction to take from the location pointed to on the map to go to another location, also pointed to on the map, where the data is obtained via the keyboard.

By using a security ink, which has no response in the visible range, the indexes can be printed on the map without impairing legibility or adversely affecting location. These advantages are of little interest in the case of homogenous areas on the map, such as areas of sea. In these spaces, identified by a cyan frame, the indexes can be printed, repetitively or otherwise, using markings provided on the cyan frame, in place of those inscribed in security ink, or in addition thereto, for example to provide specifically nautical indications, such as bathymetrical information. More generally, on a map having a background formed by a photographic image, it suffices for the frame to take the form of a simple photographic insertion in a shade which, if it is visible, interferes as little as possible with the viewing of the image.

The maps required for implementing the invention can be manufactured using a wide variety of processes, from which a choice will be made according to the application concerned.

These applications can be placed in three broad categories.

For professional applications, such as the preparation of an assignment, which generally necessitate a map provided with information proper to each particular assignment, the map can be prepared immediately prior to the assignment, with the help of an ink-jet printer, for example, using information required for the assignment retrieved from a data base, this data base being stored in a semi-conductor memory.

The map can either be produced by inscribing the indexes, using a security ink-jet printer, on an existing map, or made up in a simplified form by retrieving data from the existing map (constituted on the basis of data such as that provided by the National Geographical Society, the Naval Hydrographic Service, the National Forestry Department, etc.), edited with the location frame using an ink-jet printer in several passes.

For semi-professional applications, for example for navigators, the data does not need to be prepared for a particular assignment but distributed according to geographical areas. In this case, the map simply needs to be provided with indexes. The associated data processing means will form, in particular, the equivalent of nautical instructions, almanachs providing information on the tides, etc. A link-up with a global positioning system, or GPS, is of particular interest in this application.

Finally, the popular applications are all connected with touring and transport: they will enable the publishers of maps, guides and plans to substitute data processing means for books and, what is more, they will make it possible to reduce the diversity of maps and to develop the variety of information on a theme by theme basis, for example by providing data bases devoted to road traffic, historic monuments, points of sale, hotels, etc.

In this case, the maps can be manufactured using entirely conventional methods, except that editing necessitates a fifth offset film, this being required for printing the indexes in security ink.

One full sized example of a framed map according to the invention is shown in FIG. 5, which shows a land register map. The location frame, which is superimposed on the cartographic markings, formed in particular by streets, the contours of plots of land, and houses, is composed of indexes 10 disposed checkerboard fashion and uniformly covering the entire map, each one appearing, to the naked eye, as a single dot.