CN101082990A - Digital map shape vector encoding method and position information transfer method - Google Patents

Abstract

It is an object to provide a method of coding the position information of a digital map in a small data volume by utilizing a compressing and coding technique. In a coding method of coding data indicative of a shape vector on a digital map, an arithmetic processing is carried out over position information about a node string and a shape which indicate the shape vector to be represented by data having a statistical deviation, and the data having the statistical deviation are coded to reduce a data volume. It is possible to considerably decrease a transmission data volume in the case in which the vector shape of the digital map is to be transferred.

Description

Shape data coding-decoding method and device and positional information receiving/transmission method and device

The present invention is to be on April 26th, 2002 applying date, and application number is dividing an application of 02811758.1 application for a patent for invention " digital map shape vector encoding method and position information transfer method ".

Technical field

The present invention relates to send the coding method and the equipment thereof of the data volume of the method for positional information of numerical map and the data that coding will send, relate in particular to the shape data coding method and coding/decoding method and the device thereof that utilize compression and coding techniques minimizing data volume, and positional information reception and sending method and device thereof.

Background technology

For digital map database is installed transport information is being offered under the situation of guiding aircraft carrier aircraft device, traditionally, by connecting the wire size particular link and specifying such as the such node in crossroad that appears on the road by node number, and, even preserve the not numerical map of homology establishment with transmit leg and receiver, also can accurately send the mode of the position on the numerical map, send point on the road with respect to the method for the distance of node with expression.

But, be defined in the road net node number with connect wire size to be changed with road new replenish or change corresponding new digit.In addition, the digital map data that becomes each company in establishment source also will correspondingly obtain upgrading.Therefore utilize node number and the method that connects wire size to need a large amount of social cost to keep.

In order to improve this on the one hand, the present inventor has proposed following map position information transfer in JP-A-11-214068 and JP-A-11-243266.

In such method, when information providing will send the position of the road that causes the incident such as traffic congestion or traffic hazard, send to the take over party and to constitute the node be arranged on the road that is certain shape in the roadway segment that is comprising event location and interpolated point (if do not limit in this manual with predetermined length, be called " node " that comprise interpolated point, approach the summit of the broken line of road curve) " the road shape data " of coordinate string, with indication " event location data " based on the event location of relative position in the roadway segment of road shape data representative, that side who receives these information utilizes the road shape data to specify roadway segment on self numerical map, thereby utilizes the incident occurrence positions in the segmentation of event location data particular link.

Figure 43 has shown that " road shape data " and Figure 44 have shown " event location data ".

But, in the method for the positional information of utilizing " road shape data " and " event location data " transmission numerical map, exist the problem that data volume constantly increases and the data traffic volume also increases thereupon of the road shape data of particular link shape.

As the method for the data volume that reduces the road shape data, the present inventor has proposed to utilize splines to approach the method for the shape of road in JP-A-2001-12127.For the location revision method for sending information, be necessary further to impel data volume to reduce.

The invention solves such problem, the purpose of this invention is to provide by utilizing compression and coding techniques, with small data quantity send the positional information of numerical map position information transfer method, reduce the coding method of data volume and carry out the equipment of these methods.

Summary of the invention

The invention provides coding and represent the coding method of the data of the shape vector on the numerical map, comprise step: the positional information to each node of the relevant node strings of representing shape vector is carried out arithmetic processing, positional information is converted to the data that have statistic bias, and coded data, to reduce data volume.

The present invention also provides coding to represent the shape vector data-encoding scheme of the data of the shape vector on the numerical map, comprises step: to the shape vector that is included in numerical map in node or interpolated point location information related carry out arithmetic processing; Positional information is converted to the data that have statistic bias; And coded data is to reduce data volume.

In addition, the invention provides a kind of position information transfer method that is used for numerical map, in the method, transmit leg sends the shape data of representing the shape vector on the numerical map, and the take over party carries out map match according to the shape data that receives and on self numerical map the designated shape vector, described method comprises step: transmit leg sends the shape vector data of encoding by above-mentioned coding method; With the decode data that receive and reproduce shape of take over party, and specify and reproduce the corresponding shape vector of shape by map match.

And, the transmitter that sends the shape data of representing the shape vector on the numerical map to the take over party comprises the code table calculation element, be used for the positional information of each node of the node strings of representing the shape vector on the numerical map is carried out arithmetic processing, positional information is converted to the data that have statistic bias, distribute with appearance, generate the code table that coding is used for data according to data; With the positional information conversion equipment, be used to utilize code table.Coding to send to the take over party shape vector each node positional information and be used to generate the shape data that will send to the take over party.

In addition, the receiver that receives the coded data represent shape vector on the numerical map from transmit leg comprises the code data decoding device, be used for the reception data of decoding and coding and on numerical map reproduction by the shape data of positional information representative; And map matching means, be used to utilize the shape data of reproduction like this to carry out map match, thus on self numerical map the designated shape vector.

The present invention also provides a kind of method that representative is used for arriving by the shape data of the shape vector on the numerical map of map match designated shape the take over party that sends, comprise step:, thereby node is set with resample shape vector on the numerical map of regular length; Position data to the node that will convert the data that have statistic bias to is carried out arithmetic processing; Distribute with the appearance according to the data of storing statistic bias, generation will be used for the code table of coded data.

The present invention also provides a kind of take over party is arrived in representative by the shape data of the shape vector on the numerical map of map match designated shape method that is used to send, comprise step:, thereby node is set with resample shape vector on the numerical map of regular length; Position data to the node that will convert the data that have statistic bias to is carried out arithmetic processing; With will exist the data-switching of statistic bias to become coded representation by the identifying code table.

The present invention also provides a kind of being used for to represent the shape data of the shape vector on the numerical map and carry out the method that map match is come designated shape vector on self numerical map from the transmit leg reception, comprises step: by the reception data of identifying code tabulation sign indicating number coding; From decoded data, reproduce the shape data of the positional information representative of relevant numerical map; With utilize the shape data that so reproduces to carry out map match, thereby on self numerical map the designated shape vector.

The present invention also provides the shape data coding method of the data of the shape that coded representation represents geological information, the method comprising the steps of: utilize node strings to represent described shape, wherein, each node in the described node strings is by having at least two positions of elements information representations; Conversion about an element in the described element of the described positional information of each node of described node strings providing, so that describedly concentrate about predetermined value greater than other elements in the described element of described positional information through data converted through data converted; And it is described through data converted to encode.

The present invention also is provided for the position information transfer method of numerical map, in the method, transmit leg sends the shape data of representing the shape on this numerical map, and the take over party is mated according to received shape data and specify described shape on self numerical map, wherein, described transmit leg sends the shape data by the coding method coding of technique scheme, and decode received data and reproduce shape of described take over party, and specify and the corresponding shape of being reproduced of shape by described coupling.

The present invention also provides to the transmitter of take over party's transmission through the shape data of above-mentioned coding.

The present invention also provides the receiver that receives the shape data of encoding through said method from transmitter.

The present invention also provides the shape data coding method of the data of the shape that coded representation represents geological information, the method comprising the steps of: utilize node strings to represent described shape, wherein, each node in the described node strings is by having at least the first element and the second positions of elements information representation; For each node in a plurality of nodes: described second element of predicting the described position of the described node in the described node on the data of described shape, and obtain the difference between second element of being predicted of reality second element of described positional information of the described node in the described node and described positional information, wherein, the difference that is produced is concentrated about the predetermined value greater than this second element of these a plurality of nodes; And the differential coding that is obtained become coded data.

The present invention also provides coding and sends the transmitting apparatus of data that the shape of geological information is represented in expression, comprise: be used to utilize node strings to represent the device of described shape, wherein, each node in the described node strings is by having at least two positions of elements information representations; Be used for changing a element about the described element of the described positional information of each node of described node strings to provide, so that described through the device of data converted about concentrating greater than the predetermined value of other elements in the described element of described positional information through data converted; Described device through data converted is used to encode; And the device that is used to send described coded data.

The present invention also provides coding and sends the method for data that the shape of geological information is represented in expression, and the method comprising the steps of: utilize node strings to represent described shape, wherein, each node in the described node strings is by having at least two positions of elements information representations; Conversion about an element in the described element of the described positional information of each node of described node strings providing, so that describedly concentrate about predetermined value greater than other elements in the described element of described positional information through data converted through data converted; It is described through data converted to encode; And send described coded data.

The present invention also is provided for the receiving equipment of decoding and coding data, wherein, this coded data obtains by following steps: utilize node strings to represent to represent the shape of geological information, each node in the wherein said node strings is by having at least two positions of elements information representations; Conversion about an element in the described element of the described positional information of each node of described node strings providing, so that describedly concentrate about predetermined value greater than other elements in the described element of described positional information through data converted through data converted; And it is described through data converted to encode, and this receiving equipment comprises: the device that is used to receive this coded data; This coded data that is used to decode obtains described device through data converted; And be used to change and describedly obtain device about the described positional information of each node of described node strings through data converted.

The present invention also is provided for the method for decoding and coding data, wherein, this coded data obtains by following steps: utilize node strings to represent to represent the shape of geological information, each node in the wherein said node strings is by having at least two positions of elements information representations; Conversion about an element in the described element of the described positional information of each node of described node strings providing, so that describedly concentrate about predetermined value greater than other elements in the described element of described positional information through data converted through data converted; And it is described through data converted to encode, and the method comprising the steps of: receive this coded data; This coded data of decoding obtains described through data converted; And change and describedly obtain described positional information about each node of described node strings through data converted.

The present invention also provides coding and sends the transmitting apparatus of data that the shape of geological information is represented in expression, comprise: utilize node strings to represent the device of described shape, wherein, each node in the described node strings is by having at least the first element and the second positions of elements information representation; Be used for each node for a plurality of nodes, predict described second element of the described position of the described node in the described node on the data of described shape, and obtain the device of the difference between second element of being predicted of reality second element of described positional information of the described node in the described node and described positional information, wherein, the difference that is produced is concentrated about the predetermined value greater than this second element of these a plurality of nodes; The differential coding that is used for being obtained becomes the device of coded data; And the device that is used to send described coded data.

The present invention also provides coding and sends the method for data that the shape of geological information is represented in expression, and comprising: utilize node strings to represent described shape, wherein, each node in the described node strings is by having at least the first element and the second positions of elements information representation; For each node in a plurality of nodes: described second element of predicting the described position of the described node in the described node on the data of described shape, and obtain the difference between second element of being predicted of reality second element of described positional information of the described node in the described node and described positional information, wherein, the difference that is produced is concentrated about the predetermined value greater than this second element of these a plurality of nodes; The differential coding that is obtained is become coded data; And send described coded data.

The present invention also is provided for the receiving equipment of decoding and coding data, wherein, this coded data obtains by following steps: utilize node strings to represent shape, wherein, each node in the described node strings is by having at least the first element and the second positions of elements information representation; For each node in a plurality of nodes, predict described second element of the described position of the described node in the described node on the data of described shape, and obtain the difference between second element of being predicted of reality second element of described positional information of the described node in the described node and described positional information, wherein, the difference that is produced is concentrated about the predetermined value greater than this second element of these a plurality of nodes; And the differential coding that is obtained become coded data, this receiving equipment comprises: the device that is used to receive described coded data; The described coded data that is used to decode obtains described device through data converted; And be used for changing and describedly obtain device about the described positional information of each node of described node strings through data converted.

The present invention also is provided for the method for decoding and coding data, and wherein, this coded data obtains by following steps: utilize node strings to represent shape, wherein, each node in the described node strings is by having at least the first element and the second positions of elements information representation; For each node in a plurality of nodes, predict described second element of the described position of the described node in the described node on the data of described shape, and obtain the difference between second element of being predicted of reality second element of described positional information of the described node in the described node and described positional information, wherein, the difference that is produced is concentrated about the predetermined value greater than this second element of these a plurality of nodes; And the differential coding that is obtained become coded data, the method comprising the steps of: receive described coded data; The described coded data of decoding obtains described through data converted; And change and describedly obtain described positional information about each node in the described node strings through data converted.

Therefore, the data volume of shape vector and in the time will sending the shape vector of numerical map in the compressed digital map can reduce the data volume that will transmit considerably effectively.The take over party, reconstruct shape data and carry out map match from receive data is so that can accurately specify the shape vector of transmission.

Description of drawings

Fig. 1 is the figure that is presented at the node of resampling under the situation of application based on the coding method of the 1st embodiment;

Fig. 2 is based on the code table in the coding method of the 1st embodiment;

Fig. 3 is the run length code table that will be used in based in the coding method of the 2nd embodiment;

Fig. 4 is the code table that will be used in based on the relevant Δ θ in the coding method of the 2nd embodiment;

Fig. 5 is the code table that will be used in based on the relevant Δ θ that run length is taken in the coding method of the 2nd embodiment;

Fig. 6 shows the calcspar of execution based on the structure of the equipment of the position information transfer method of the 3rd embodiment;

Fig. 7 is the process flow diagram that is presented at based on the code table constructive process in the coding method of the 3rd embodiment;

Fig. 8 is the process flow diagram that is presented at based on the shape data constructive process in the coding method of the 3rd embodiment;

Fig. 9 is in the position information transfer method that is presented at based on the 3rd embodiment, road/segmentation is appointed as the form of the structure of the shape vector serial data information that sends data;

Figure 10 is in the position information transfer method that is presented at based on the 3rd embodiment, relevant as the shape vector data that send data, with respect to the form of the structure of the various transport information of the relative distance representative of each node;

Figure 11 is in the position information transfer method that is presented at based on the 3rd embodiment, at the process flow diagram of take over party's processing procedure;

Figure 12 is in the coding method that is presented at based on the 4th embodiment, the figure of the relation between the curvature of sampling section length and shape data;

Figure 13 is in the coding method based on the 4th embodiment, and the figure that circular arc and straight line are approximate is described;

Figure 14 is the figure of the segmentation of dividing in the coding method that is presented at based on the 4th embodiment;

Figure 15 (a) is presented in the coding method based on the 4th embodiment, the process flow diagram of the method for the section length of determining to resample;

Figure 15 (b) is the figure of the form quoted in the process flow diagram that is presented at shown in Figure 15 (a);

Figure 16 is in the coding method based on the 4th embodiment, and the figure that quantification is resampled is described;

Figure 17 is in the coding method based on the 4th embodiment, and the figure of the candidate point of next node is described;

Figure 18 is the process flow diagram that is presented at based on the node deterministic process in the coding method of the 4th embodiment;

Figure 19 is based on the code table in the coding method of the 4th embodiment;

Figure 20 is the process flow diagram that is presented at based on the code table constructive process in the coding method of the 4th embodiment;

Figure 21 is the process flow diagram that is presented at based on the shape data constructive process in the coding method of the 4th embodiment;

Figure 22 is the form that is presented at based on sending the structure of data in the position information transfer method of the 4th embodiment;

Figure 23 (a) and (b) and (c) be the figure that typically is presented at based on the transmission of data in the coding method of the 4th embodiment;

Figure 24 be presented at based in the position information transfer method of the 4th embodiment at the process flow diagram of take over party's processing procedure;

Figure 25 is the figure that shows based on the applied node location of the coding method of the 5th embodiment, distance and angle information;

Figure 26 (a) and (b) are the code tables that will be used in based in the coding method of the 5th embodiment;

Figure 27 is the process flow diagram that shows based on the code table constructive process in the coding method of the 5th embodiment;

Figure 28 is the process flow diagram that shows based on the shape data constructive process in the coding method of the 5th embodiment;

Figure 29 is in the position information transfer method that is presented at based on the 5th embodiment, road/segmentation is appointed as the form of the structure of the shape vector serial data information that sends data;

Figure 30 is in the position information transfer method that is presented at based on the 5th embodiment, relevant shape vector data, with respect to the form of the structure of the various transport information of the relative distance representative of each node;

Figure 31 is presented at the figure of application based on the node location under the situation of the coding method of the 6th embodiment, distance and angle information;

Figure 32 is the code table that will be used in based in the coding method of the 6th embodiment;

Figure 33 is the process flow diagram that shows based on the code table constructive process in the coding method of the 6th embodiment;

Figure 34 is the process flow diagram that shows based on the shape data constructive process in the coding method of the 6th embodiment;

Figure 35 is in the position information transfer method that is presented at based on the 6th embodiment, road/segmentation is appointed as the form of the structure of the shape vector serial data information that sends data;

Figure 36 shows the view be suitable for using based on the shape of the road of the coding method of the 7th embodiment;

Figure 37 is the process flow diagram that shows based on the θ code table constructive process in the coding method of the 7th embodiment;

Figure 38 is the process flow diagram that shows based on the Δ θ code table constructive process in the coding method of the 7th embodiment;

Figure 39 is the process flow diagram that shows based on the shape data code table constructive process in the coding method of the 7th embodiment;

Figure 40 is in the position information transfer method that is presented at based on the 7th embodiment, road/segmentation is appointed as the form of the structure of the shape vector serial data information that sends data;

Figure 41 is the figure that is used to illustrate the distance and the angle of specified coordinate point;

Figure 42 (a) and (a ') are the figure and the curve maps of the total curvature function representation of display shape data;

Figure 42 (b) and (b ') are figure and the curve maps that the angle of deviation of display shape data is represented;

Figure 42 (c) and (c ') are figure and the curve maps that the prediction value difference of the angle of deviation of display shape data is represented;

Figure 43 is the form that shows the data structure of the shape vector serial data information in the traditional location method for sending information; With

Figure 44 is the form that shows the data structure of the transport information in the traditional location method for sending information.

In addition, in the label in the drawings, the online processing section of 10 and 30 expressions, 11 presentation of events information input unit branches, 12 expression numerical map display parts, 13 and 22 expression digital map databases, 14 expression map match parts, 15 expression positional information conversion portions, 16 expression positional informations send part, 17 expression positional information receiving units, 18 expression code data decompression parts, 20 expression processed offline parts, 21 expressions are transport information in the past, 23 expression code table calculating sections, 24 expression code table data and 40 expression roads.

Preferred embodiment describes in detail

(the 1st embodiment)

In the 1st embodiment, will be to being described by the variable length code method of compressing data.

In the method for the positional information that sends numerical map according to the present invention, at first, the shape of road is with existing the shape data of statistic bias to represent.Its reason is that compressibility will be improved when compression and coded data.

In the shape of road with being arranged under the situation that the coordinate points on the road represents, as shown in figure 41, by with respect to adjacent coordinates point (P _J-1) distance and this two dimension of angle can specify each coordinate points (P uniquely _J) the position.In Figure 41, this angle is based on is arranged to 0 and specify the angle Θ of " absolute azimuth " of certain angle between 0 to 360 along clockwise direction to (making progress an in the drawings) position angle, positive north _jTherefore, be called as the total curvature function representation with distance and absolute azimuth denotation coordination point.

When with coordinate points P _J-1, P _JAnd P _J+1X and y coordinate be set to (x _J-1, y _J-1), (x _j, y _j) and (x _J+1, y _J+1) time, distance L _j(coordinate points P _JAnd P _J+1Between distance) and absolute angle Θ _j(from coordinate points P _JExtend to coordinate points P _J+1The absolute azimuth of straight line) can calculate by following equation:

$L_j=\sqrt{\{{(x_{j+1}-x_j)}^2+{(y_{j+1}-y_j)}^2\}}$

Θ _j＝tan ^-1{(x _j+1-x _j)/(y _j+1-y _j)}

Resetting (resampling) new coordinate points, so that be that constant (under=L) the situation, at the vpg connection of road, only sends relevant angle Θ for each coordinate points with respect to the distance of adjacent coordinates point _jInformation (that is, one-dimension information) so that the take over party can specified coordinate point the position, reduce the data volume that except the common information of relevant L, will send.

Figure 42 (a) has shown the coordinate points of resampling in such position, so as with respect to the distance of the adjacent coordinates point on the road be constant (under=L) the situation, each coordinate points (P _J) absolute azimuth Θ _jAt each coordinate points absolute azimuth Θ _jUnder the situation of expression, indicate the angle information Θ of each coordinate points _jThe frequency of occurrences do not have statistic bias shown in Figure 42 (a ').

But the angle of each coordinate points also can be used the difference that replaces absolute azimuth, that is, and and " angle of deviation " θ shown in Figure 42 (b) _jRepresent." angle of deviation " θ _jCan be by following calculating.

θ _j＝Θ _j-Θ _j-1

At each coordinate points angle of deviation θ _jUnder the situation of expression, in the zone that has a large amount of straight line roads, shown in Figure 42 (b '), indicate the angle of deviation θ of each coordinate points _jThe frequency of occurrences ° locate to occur maximal value in θ=0.

In addition, shown in Figure 42 (c), the angle of each coordinate points also can be used angle of deviation θ _jWith angle of deviation statistical forecast value S _jDifference DELTA θ between (predicted value of representing with angle of deviation) _jExpression.Angle of deviation statistical forecast value S _jBe by utilizing up to P _J-1The angle of deviation estimated coordinates point P of former coordinate points _JAngle of deviation θ _jObtain.For example, can be angle of deviation statistical forecast value S _jBe defined as:

S _j＝θ _j-1

Or

S _j＝(θ _j-1+θ _j-2)/2

In addition, angle of deviation statistical forecast value S _jAlso can be by the weighted mean of the angle of deviation of past n coordinate points is arranged to S _jDefine.The prediction value difference of angle of deviation is pressed following calculating:

Δθ _j＝θ _j-S _j

Most road shape is straight line or slow sweep.Therefore, be configured under the situation of constant the prediction value difference Δ θ of angle of deviation in the distance L between the coordinate points _jConcentrate near 0 °, make shown in Figure 42 (c ') the angle information Δ θ of each coordinate points _jThe frequency of occurrences near θ=0 °, have very big deviation.

In order to obtain to exist the shape data of statistic bias, to have the section length L that resamples of constant distance, the road shape (original-shape) of taking a sample at regular intervals, sampling spot (node) P _JPosition data angle of deviation θ _jPrediction value difference Δ θ _j(=θ _j-S _j) expression.Distance can be by being generalized to the extraneous actual range that obtains or being the length of unit representation with predetermined normalization coordinate.

Here stipulate, angle of deviation statistical forecast angle S _jBe made as

S _j＝(θ _j-1+θ _j-2)/2

Because in most of the cases, the shape of road is slowly crooked,

θ _j≈(θ _j-1+θ _j-2)/2＝S _j

Therefore, can suppose Δ θ _jBe distributed near the scope very little 0.

In theory, Δ θ _jCan have the value between-360 ° to+360 °.For this reason, for representing that resolution is 1 ° Δ θ _j, need will represent 1 of sign to add 10 positions that together obtain with 9 positions representing 360 numerical value.Be coded in ± near angle 0 ° and by using, be used for coded delta θ the angle that is assigned to than 10 big values in position away from ± 0 ° of setting than 10 little values in position _jAverage figure place can be configured to less than 10 positions and shape data is whole can represent with little data volume.

Fig. 2 has shown that the code that will be used to encode is assigned to the code table of Δ θ.If Δ θ=0 is set, so, proceed to zero coding.If Δ θ=+ 1 is set, so, the additional bit of representing positive sign (overheadbit) 0 is added in the code 100, obtain 1000.If Δ θ=-1 is set, so, the additional bit 1 of representing negative sign is added in the code 100, obtain 1001.

Referring now to Fig. 1 variable length code is described.If the node number is 6 (=beginning edge+5 node), except angle initial value (10 positions), normal encoding need have the fixed-length data quantity of a position, 5 * 10 position=50.Otherwise, in the situation of utilizing coding schedule coding as shown in Figure 2, if hypothesis Δ θ _j Get 0 value 3 times and equal less than ± 2 ° value 2 times, so, except angle initial value (10 positions), data volume can be represented with a position, 3 * 1 position+2 * 4=11 positions.If data are " 0,0 ,+1 ,-2,0 ", can represent them with " 00100010110 " by coding.

The take over party can use the value of Δ θ successively with reference to the code table that sends (or preserving in advance) with shape data, obtains Δ θ _jEach value.By beginning to add up successively, can determine the angle of deviation θ of each coordinate points uniquely from initial value _jValue.

Code table is by calculating each coordinate points P _JΔ θ _jAngle is checked what the frequency of occurrences of angle and the well-known Hofman tree structure that utilization depends on the frequency of occurrences were created.

Therefore, the shape data that has the coding deviation is carried out arithmetic processing, then, carry out variable length code.Like this, can reduce the data volume of shape data.

Node location is represented with distance between the adjacent node and angle of deviation though resample,, section length L sampling node location at regular intervals also can be with relative latitude and longitude coordinate (Δ x to resample _j, Δ y _j) represent.In this case, according to S _Jx=Δ x _J-1And S _Jy=Δ y _J-1Definition, statistical value S _jRepresent by following formula:

Δx _j＝S _jx+δx _j＝Δx _j-1+δx _j

Δy _j＝S _jy+δy _j＝Δy _j-1+δy _j，

And, δ x _jWith δ y _jBe variable length code, therefore, can be used as shape data and send.

(the 2nd embodiment)

In the 2nd embodiment, described utilizing run length method method of compressing data.

In the example of the 1st embodiment, Δ θ _jBe encoded under the situation of expression shape data, in the crooked in the same manner road of straight line road or curvature, " 0 " is repetition.Under these circumstances, data compressibility ratio in the expression formula of " 0 repeats 20 times " is taller in " 00000... ".Here, carry out run length coding, RLC and come packed data.

Fig. 3 has shown definition, and for example, same number repeats the code table that 5 times (run length is 5) is shown as the relevant run length of " 101 ".Fig. 4 has shown the code table of the relevant Δ θ identical with code table among Fig. 2.

For example, data array is defined as run length → Δ θ → run length → Δ θ → ...When Δ θ is " 0,0,0,0,0 ,-2 ,-2,0 ,+3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 ,-1.. ",

Can represent with " 101.0_0.1011_0.1011_0.0_0.11000_1101.0_0.1001.. " → " 10100101101011000110001101001001... " (32 positions) by run length coding, RLC.

On the other hand, under the situation of not using run length to represent, " 0000010111011011000000 00000000000001001... " (38 positions) is set.

In addition, the code table shown in Fig. 3 and 4 can be limited in also can comprehensively being become a kind of code table as shown in Figure 5 on the effective especially form.In Fig. 5, only at the situation of Δ θ=0 run length of giving a definition.By utilizing the code table among Fig. 5, " 0,0,0,0; 0 ,-2 ,-2,0 ,+3; 0,0,0,0,0; 0,0,0,0; 0,0,0,0; 0,0,0 ,-1.. " can be represented with " 100.11101.11101.0.111100.1100.11011... " → " 10011101111010111100110011011... " (29 positions).

In this connection, compare, obtain 10 position * 25 nodes=250 positions with method for expressing for the regular length of 10 positions of 1 node.Obviously, in this example, be compressed into 29 ÷ 250=12%.

(the 3rd embodiment)

In the 3rd example, the equipment of the executing location method for sending information according to the present invention is described.

Fig. 6 has shown an example of the equipment of having done, with the positional information transmitter/receiver of event occurrence information on another equipment 30 exchange roads.

This equipment comprises the processed offline part, is used under off-line state generating will being used to compress and the code table of the road shape data of encoding; With online treatment part 10, the code table data that are used to utilize processed offline part 20 to generate send transport information.Processed offline part 20 comprises digital map database 22; Storage area 21 is used for storage transport information in the past; Code table calculating section 23 is used to generate and will be used to the code table data compressing and encode; With the code table database, be used to store the code table data of generation like this.

On the other hand, online treatment part 10 comprises positional information receiving unit 17, and the positional information that is used for slave unit 30 sends " the road shape data " and " event location data " that part 16 receives through overcompression and coding; Code data decompression part 18, (decoding) compression and coded data are used for decompressing; Digital map database 13 is used to store digital map data; Map match part 14 is used to utilize road shape data and event location data through decompressing to carry out map match and be used for allocate event position on self numerical map; Numerical map display part 12 is used for stack and presented event position on map; Event information importation 11 is used to import the information of relevant generation incident; Positional information conversion portion 15, be used to determine to comprise the target road segmentation of event location, generate " the event location data " represent event location, as the relative position of target road segmentation with utilize the shape data of 24 compressions of code table data and the target road segmentation of encoding; Send part 16 with positional information, be used for " the road shape data " and " event location data " that generate are sent to the positional information receiving unit 17 of equipment 30.

Process flow diagram among Fig. 7 and Fig. 8 has shown the process of operation of equipment.In processed offline part 20, code table calculating section 23 exists

Step 1: with reference to past transport information 21;

Step 2: select the target road segmentation of transport information; With

Step 3: with the resample shape data of target road segmentation of regular length L.

Step 4: the position data of node is converted to the total curvature function representation.

Step 5: the Δ θ that calculates each segmentation/each node according to the statistical value accounting equation.

Step 6: then, the appearance of calculating Δ θ distributes.

Step 7: then, calculate the continuous distribution of same value.

Step 8: create code table according to the appearance distribution of Δ θ and the continuous distribution of same value.

Step 9: the code table of finishing is stored in the code table database 24.

This processing procedure defines by the program that the computing machine that makes processed offline part 20 plays 23 effects of code table calculating section.

In addition, in online treatment part 10, positional information conversion portion 15 exists

Step 10: divide 11 receiving traffic informations from the incident information input unit;

Step 11: selection comprises the target road segmentation of the occurrence positions of traffic events; With

Step 12:, thereby as shown in Figure 8 node is set with the resample shape data of target road segmentation of regular length L.

Step 13: the position data of node is converted to the total curvature function representation.

Step 14: the Δ θ that calculates each segmentation/each node according to the statistical value accounting equation.

Step 15:, shape data is converted to coded representation by with reference to the code table data 24 that are intended to the code table that the target road segmentation creates (or be intended to have near the link creation of the shape of target road segmentation code table).

Step 16: send shape data through the relevant target road segmentation of coding with the data of relevant event location with the relevant relative information representative of target road segmentation.

This processing procedure defines by the program that the computing machine that makes online treatment part 10 plays 15 effects of positional information conversion portion.

Fig. 9 and Figure 10 have shown the road shape data (Fig. 9) and the event location data (Figure 10) that will send.The data of the relevant section length L that the road shape data comprise the code table data, resample and through the shape data of overcompression and coding.

Figure 11 has shown the processing procedure that receives the take over party of data.

Step 20: when positional information receiving unit 17 receiving position informations,

Step 21: code data decompression part 18 is by with reference to being included in the code table reconfiguration code data that receive in the data and shape data being converted to the total curvature function.

Step 22: then, reproduce the shape data of representing with latitude and longitude coordinate.

Step 23: the road shape that 14 pairs of map match parts are reproduced shape and self numerical map carries out map match, with the intended target roadway segment, and according to the event location data, specifies the traffic events occurrence positions in the target road segmentation.

Step 24: numerical map display part 12 superposes on map and shows transport information.

This processing procedure defines by the program that the computing machine that makes online treatment part 10 plays code data decompression part 18 and 14 effects of map match part.

Though the code table that is used in compression and the coding is included in the data that will send and therefore obtains to send,, have identical code table by making to send in advance with the take over party, code table is included in the data that will send.

We have described the code table data of being created by processed offline part 20 24 and have been used to obtain example by the shape data of 10 compressions of online treatment part and coding.Each road shape of 20 compressions of processed offline part and coding target area and preservation in advance are by the shape data of each roadway segment of coded representation.When the information of the generation that obtains relevant traffic events, online treatment part 10 can select to comprise the road shape data of encoding in the roadway segment of traffic events occurrence positions from the shape data being kept at processed offline part 20, generation is represented the transport information of traffic events occurrence positions by the relative position of roadway segment, and sends the coding road shape data of selection like this and the transport information of generation to the take over party.

In this case, processed offline part 20 is in the process of step 2 to 9, and resampling with regular length L to become the shape data of roadway segment of coding target, calculates the Δ θ of each node and distributes according to the appearance of Δ θ, creates code table.Then,, the Δ θ of relevant each coordinate points of resampling is converted to coded representation, create the shape data of compression and coding, it is stored in the database by using the code table of establishment like this.Repeat to handle by each roadway segment, can preserve the compression and the coding shape data of each roadway segment that is included in the target area the target area.

Therefore, in online treatment, also can utilize under off-line state the result who resamples who road shape is carried out with regular length L.In online treatment, also can utilize under off-line state the result who resamples who road shape is carried out with regular length L.

In the 4th embodiment, the method that irreversible compression method is introduced in the high compression road shape data is described.

In the process of transmitting of speech data and view data, carry out aspect sensation (vision/sense of hearing), reducing the processing of the sampling spot in the adjacent scope, so that the raising compressibility reduces the quantization digit of metrical information or the processing of management metrical information, so that improve compressibility.Under the situation of carrying out these processing, can not fully reconstruct raw data the take over party.Even more or less change in data, do not have under the situation of what inconvenience yet, handle by introducing irreversible compression, considerably packed data.

In based on position information transfer method of the present invention, the take over party carries out map match and comes the particular link shape.Therefore, be necessary to send the accurate shape that the starting point and the terminal point of target road segmentation and being easy to causes the part of mismatch.In other parts,, also can specify the original position the take over party even the shape that will send is somewhat smudgy.Therefore, in based on position information transfer method of the present invention, can also introduce the compressibility that irreversible compression method improves the data that will send.

In compression method, improve the compressibility of data by the following method based on present embodiment.

(1) reduces the interior sampling spot of scope that does not cause mismatch.On the very big height crankcase ventilaton of curvature, match point skew road may cause mismatch.For this reason, according to as shown in figure 12 curvature sampling section length L is set.

(2) reduce the quantization digit of representing the Δ θ in the scope that does not cause mismatch.For example, minimum resolution being arranged to 2 °, is that unit quantizes Δ θ with this value.In this case, near true value, laterally move and reproduce node location, so that the reproduction shape is deformed in the nature of things.For this reason, the take over party carries out the interpolation processing of smooth reproduction shape.

(3) make the shape approximation circular arc and the straight line of road.Have deviation in the value of making in based on the poor expression formula of angle of deviation coded prediction value Sj, so as to compress and the situation of encoding under, deviation concentrates on above freezing in the identical roadway segment of the curvature of representing with circular arc or straight line.Therefore, further improve statistic bias, so that improve compression efficiency rapidly.In view of the above, make the shape approximation circular arc and the straight line of road, so that can improve compressibility rapidly.In addition, can also improve the effect of run length.

Method (1), (2) and (3) can be carried out separately, also can combine execution.

Now the special case of compressing by application process (1), (2) and (3) and encode is described.

The shape approximation circular arc and the straight line of road can be realized by the shape of linear-apporximation with the road of total curvature function representation.As shown in figure 13, represent θ (=∑ θ at axis of ordinates _j) and abscissa axis represent L (=∑ L _i) (L _iBe constant) coordinate system in, use the some P on the total curvature function representation road as shown in Figure 1 _JThe shape of road be shown as solid-line curve.Make the straight line that the shape approximation circular arc of road and straight line represent to make curve approximation shown in dotted line (θ=aL+b).(θ=b) represents the linearity configuration of road and the straight line of slope ≠ 0, and (θ=aL+b) represents the arcuate shape of road to the straight line of slope in coordinate system=0.

In this being similar to,, determine allowable error along roadway segment by the method (JP-A-2001-129665 and JP-A-2001-132611) that the inventor separately proposes.In the method, the tolerance limit (allowing azimuth angle error) of the tolerance limit of the error of relevant distance (distance error) (allowable distance error) and relevant azimuthal error (azimuth angle error) is arranged to allowable error, so that be that unit satisfies following condition with each node in the shape that is included in road or every line.

1. near the starting point and terminal point of target road segmentation, be provided with the allowable distance mistake a little bit smallerly.

2. under parallel road situation adjacent one another are, be provided with the allowable distance mistake a little bit smallerly.

3. near the crossroad that has the linking-up road that the angle of the crossing is little such as switching access or outlet, be provided with the allowable distance mistake a little bit smallerly.

If 4. shorter with respect to the distance of road on every side, allowing that the position angle mistake is provided with lessly.

5. the possibility that the difference of azimuth angle error may increase in the very big road shape part of curvature is very high.Therefore, allowing that the position angle mistake is provided with lessly.

In addition, the allowable error that each node is set on the left side and the right side of target road segmentation respectively.In application, having specifically described with the node is the method for unit quantitative Analysis allowable error.

If determine allowable error, so, make the shape approximation circular arc and the straight line of road, so that the scope of input allowable error along roadway segment.As shown in figure 14, the shape of road is divided into the segmentation of representing with circular arc or straight line.

Then, determine each section length of resampling of segmentation.

By following equation, depend on the curvature a of each segmentation _j, for the section length of resampling is determined in each segmentation:

L _j＝K×1/|a _j|

(K is a predetermined constant)

In addition, can quantize L _jValue.If by quantizing L _jThe value of being got is, for example, and 40/80/160/320/640/1280/2560/5120 meter, so, can be L _jValue be encoded into 3 positions that will send.

At this moment, if resample section length L _jBetween adjacent sectional, do not fluctuate, can improve compression efficiency.Figure 15 (a) has been if shown that curved transition is less, and section length is defined as repetition, so that definite process of the fluctuation of the section length that suppresses to resample.According to the 1st segmentation (step 30), successively from the curvature a of each segmentation j _jIn the section length calculated value D that obtains to resample _j(step 31) obtains with respect to the section length calculated value D that resamples in the adjacent sectional _J-1Rate of change H _j(=| D _j-D _J-1|/D _j(step 32), and, the section length L that resamples in acquisition and the adjacent sectional _J-1Ratio I _j(=D _j/ L _J-1) (step 33).Rate of change H with the section length calculated value of resampling _jWith predetermined constant H _aCompare, and, will with the section length L that resamples in the adjacent sectional _J-1Ratio I _jWith predetermined value I _A1And I _A2Compare.As ratio H _jEqual less than H _aAnd I _jHas I _A1With I _A2Between value the time, the section length L that resamples _jBe arranged to equal the section length L that resamples in the adjacent sectional _J-1(step 35).Section length calculated value D will resample _jWith the section length L that resamples in the adjacent sectional _J-1The reason of comparing is that section length calculated value D resamples _jRate of change H _jLittle, thus the section length of resampling repeatedly is arranged to have identical value, to prevent the section length calculated value D that resamples _jWith the section length L that resamples _jThere are differences.

If judgement is (NO) that negates in step 34, according to D is set _jScope and the following table among the Figure 15 of the relation between the section length, from D _jValue in determine section length L _j(step 36).This processing (step 37 and 38) is carried out in all segmentations.

H _aBe arranged to have approximate 0.2 value.I _A1Be arranged to have approximate 0.7 value and I _A2Be arranged to have approximate 2.0 value.

As shown in figure 16, then, at regular intervals with the section length L that resamples _nEach segmentation n that takes a sample is to obtain node P _J, and calculate P _JAngle of deviation θ _jWith angle of deviation statistical forecast value S _jBetween prediction value difference Δ θ _j(=θ _j-S _j) quantized value.

Suppose angle of deviation θ previous node _J-1As angle of deviation statistical forecast value S _j(S _j=θ _J-1).

In addition, minimum resolution Δ θ _jBe arranged to δ ° and obtain Δ θ (minimum resolution δ) _jQuantized value.

At this moment, be that unit is provided with Δ θ with δ _jTherefore, according to distance L _nWith angle information Δ θ _jFrom previous node P _JThe node P that reproduces _J+1Always be not positioned on the original road shape (approximate shapes).As shown in figure 17, when will be from P _JObtain next node P _J+1The time, by getting Δ θ _jQuantized value, node P appears _J+1Some candidate points.In range of allowable error, from candidate point, select next node P _J+1, so that as far as possible continuously the value of Δ θ is arranged to zero.In addition, carry out such node continuously and select,, select node along the direction of the error between position that reduces selected node and the true value (at the point on the original road shape) if so that same error is increased near the limit of allowable error.In addition, in this case, select node like this, make Δ θ be arranged to zero continuously.

Figure 18 shown from a node P _J+1Relevant a plurality of candidate point P _J+1(i) select the process of a candidate point in.

Step 40: candidate point P _J+1(i) be arranged on and P _JAt a distance of L _nOn the position of Δ θ=δ i.I represents to become to have-m ... ,-1,0,1 ..., the quantized value of near zero (2m+1) the individual positive integer of m and the Δ θ of negative integer.

Step 41: calculate each candidate point P _J+1(i) and the distance D between the closest approach of original road shape _iCrossing position angle and candidate point P with the closest approach _J+1(i) the error delta Θ between the crossing position angle _i

Step 42: by following equation is each candidate point P _J+1(i) calculate estimated value ε _i

ε _i＝α·(δ·|i|)+β·D _i+γ·|ΔΘ _i|+Ψ

α, β, γ: pre-determined factor

Ψ: the penalties that is provided with during greater than range of allowable error.

Step 43: ε _iMinimum candidate point P _J+1(i) as node P _J+1

Estimated value ε _iBe that i=0 is up to making D _iWith Δ Θ _iIncrease and the minimum value till the penalties Ψ adding wherein.In view of the above, use candidate point like this, make Δ θ be arranged to zero.

In addition, about the section length D of segmentation n _nThe remnant, handle as follows.

L _n＜L _N+1: pass through distance L _nSegmentation n resamples.If the remainder of segmentation n (remnant) is less than L _n, pass through L _nResample segmentation n+1 and former segmentation are so that pass through L _N+1Distance and this point among the segmentation n+1 and the point subsequently of resampling and obtaining by a part of addition of this remnant and segmentation n+1.

L _n＞L _N+1: by the distance L n segmentation n that resamples.If the remnant of segmentation n is less than L _n, pass through L _N+1This point and segmentation n+1 among the segmentation n resample.

Therefore, by resampling, can prevent that precision from degenerating to lack section length.

If the minimum resolution δ of Δ θ ° increase, so, the expression figure place of angle reduces and the shape follow-up characteristic of circular arc degenerates, and makes the probability of Δ θ=0 reduce, thereby makes coding and compression effectiveness variation.In contrast, if δ ° of minimizing, so, the expression figure place of angle increases and the shape follow-up characteristic of circular arc is improved, and makes the probability increase of Δ θ=0 and coding and compression effectiveness also be improved.In addition, the run length compression effectiveness also is improved.Consider this aspect, be necessary minimum resolution δ ° of definite Δ θ that in fact will use.

Then, the coding of data is in this case described.

The prediction value difference Δ θ of respective nodes encodes like this, makes data length reduce about Δ θ=0.

Because most of continuous datas are data of relevant Δ θ=0, therefore, the run length of coded delta θ=0.

The section length change sign indicating number of the change point of indicating the section length of resampling is set in addition.Private code is appointed as section length sign indicating number and fixed bit (about 3 positions) the definition segment length by just in time providing after private code are provided.

Further, the reference point of identification code that reference point node in each segmentation of indication is set is provided with sign indicating number.Private code is appointed as reference point sign indicating number is set, the fixed bit that just in time provides after private code (about 6 positions) is arranged to the reference node period, with be defined as reference mode (the node number initial value be scheduled to appearing at coordinate after the reference node period, need not additional bit, and, also can use when finding code, to add 1 node number structure).

In addition, private code is appointed as EOD (ED) sign indicating number that designation data finishes.By this code, the end that the shape data string list shows is set.

Figure 19 has shown the code table that will be used to encode.

Further, Figure 20 has shown that process and Figure 21 of creating code table under off-line state have shown by utilizing code table, issue the process that communication ceases of delivering at presence.In Figure 20,

Step 50: with reference to the past transport information.

Step 51: select the target road segmentation of transport information.

Step 52: calculate range of allowable error along the target road segmentation.

Step 53: the node of target road segmentation is converted to the total curvature function representation.

Step 54: the approximate circular arc of shape vector and the straight line that make the target road segmentation.

Step 55: the length L of resampling of determining each segmentation n of approximate circular arc or straight line _n

Step 56: pass through L _nQuantize and the target road segmentation of resampling on shape data and a node is set.

Step 57: the Δ θ that calculates each segmentation/each node according to the statistical value accounting equation.

Step 58: the appearance of calculating Δ θ distributes.

Step 59: the continuous distribution of calculating same value.

Step 60: create code table according to the appearance distribution of Δ θ and the continuous distribution of same value.

Step 61: the code table of finishing is stored in the code table database 24.

In addition, carry out online treatment among Figure 21 as follows.

Step 62: divide 11 receiving traffic informations from the incident information input unit.

Step 63: selection comprises the target road segmentation of the occurrence positions of traffic events.

Step 64: calculate range of allowable error along the target road segmentation.

Step 65: the node of target road segmentation is converted to the total curvature function representation.

Step 66: the approximate circular arc of shape vector and the straight line that make the target road segmentation.

Step 67: the length L of resampling of determining each segmentation n of approximate circular arc or straight line _n

Step 68: pass through L _nQuantize and the target road segmentation of resampling on shape data and a node is set.

Step 69: the Δ θ that calculates each segmentation/each node according to the statistical value accounting equation.

Step 70: shape data is converted to coded representation by the identifying code table.

Step 71: send shape data in the coding target road segmentation with transport information.

Though we have described the example that only utilizes the data of the relevant code table of creating in online treatment in processed offline, but, also can generate and be stored in the shape data of each roadway segment of each road shape that passes through the coded representation target area in the processed offline in advance, when in online treatment, importing the information of generation of relevant traffic events with box lunch, selection comprises the coding road shape data of the roadway segment of traffic events occurrence positions from the shape data that generates processed offline, generate transport information by the relative position that represent traffic incident occurrence positions of roadway segment, with as described in the 3rd embodiment, send the coding road shape data of selection like this and the transport information that so generates to the take over party.Therefore, in online treatment, also can utilize under off-line state the result who resamples who road shape is carried out with regular length L.

Figure 22 has shown the road shape data that will send.These data comprise code table data and coding shape data, and comprise such as the reference mode of Δ θ, each segmentation and the data the sampling section length, as the coding shape data.

Figure 23 (a) and (b) and (c) typically shown the data that will between transmit leg and take over party, exchange.At transmit leg, calculate quantification and the node location afterwards of resampling, with the road shape of expression shown in Figure 23 (a), and shown in Figure 23 (b), the data of instructs node position send to the take over party.Level and smooth data that receive of take over party and the shape of reproduction shown in Figure 23 (c).In this case, can carry out interpolation (the interpolation curve such as Beziers batten or Beziers curve be suitable for) or level and smooth based on smooth function based on B batten and so on.In addition, the crossing position angle of each interpolated point of generation is evenly distributed.

Figure 24 has shown take over party's process.

Step 80: receiving position information.

Step 81: the shape data of coded representation is converted to the total curvature function by the identifying code table.

Step 82: then, proceed to the conversion and the level and smooth and interpolation processing of latitude and longitude coordinate, to reproduce shape data.

Step 83: obtain the reference mode position.

Step 84: carry out map match, with the intended target roadway segment.

Step 85: reproduce transport information.

Therefore, by using described in the present embodiment irreversible compression method, shape data obtains high compression, makes and can reduce the data volume that will send considerably.

Except as mentioned above in advance the approximate shapes, can utilize simultaneously also that quantizing resamples and carry out the circular arc and the straight line of the shape data by the total curvature function representation approximate.

The decision logic relation of the section length of resampling described above and definite process that quantizes to resample also can be applied to not make the situation of the approximate circular arc of shape data.

(the 5th embodiment)

In the 5th embodiment, the method for the coding road shape data of resampling of not utilizing coordinate points is described.

As reference Figure 41 is described, by with respect to adjacent coordinates point (P _J-1) distance and this two dimension of angle can specify the coordinate points (P that is arranged on the road uniquely _J).In the 1st to the 4th embodiment, resample like this in the position of coordinate points, and distance is remained unchanged, and the angle of only encoding is to reduce the data volume that will send.But, in this case, the processing of need resampling.

On the other hand, being used for coordinate points coding road shape data conditions to node in the road shape that is included in numerical map and interpolated point definitely, do not need the processing of resampling.In this case, the distance of node or interpolated point is not a constant.Therefore, be necessary the angle and distance of encoding.

Figure 25 has shown the method for the not only coding and decoding but also the distance of encoding.The coding of angle is identical with angular coding among the 1st embodiment, each node (comprising interpolated point) P _JAngle information be used as angle of deviation θ _jWith angle of deviation statistical forecast value S _jBetween the prediction value difference Δ θ of difference _jExpression, Δ θ _jWith, for example, 1 ° for unit (can use such as with 2 ° of other resolution as unit) is quantized, and the code table of relevant Δ θ is according to quantizing Δ θ _jOccurrence frequency create.At this moment, angle of deviation statistical forecast value S _jBe defined by, for example, S _j=θ _J-1Or S _j=(θ _J-1+ θ _J-2)/2.

Figure 26 (b) has shown the code table of the relevant Δ θ of establishment like this.This table is identical with code table (Fig. 2) based on the 1st embodiment.By using the code table of relevant Δ θ, the angle information of each node of variable length code (Δ θ _j).

On the other hand, the distance of encoding as follows.

At first, each node P _JRange information be used as adjacent node P _J+1Distance L _jWith distance statistics predicted value T _jBetween the prediction value difference Δ L of difference _j(=L _j-T _j) expression, Δ L _jWith, for example, 10m is that unit (can use such as with 50m or 100m other resolution as unit) is quantized.At this moment, distance statistics predicted value T _jBe defined by, for example, T _j=L _J-1Or T _j=(L _J-1+ L _J-2)/2.

Then, according to quantizing Δ L _jOccurrence frequency create the code table of relevant Δ L.Figure 26 (a) has shown the code table of the relevant Δ L of establishment like this.In order to represent the sign of Δ L, add the additional bit of code table.When Δ L ≠ 0 is set, if Δ L is positive, add 0 and Δ L bear, add 1.Therefore, if definition of T _j=L _J-1, so,

Work as L _jGreater than L _J-1(L _j-L _J-1＞0) time, add 0 and

Work as L _jLess than L _J-1(L _j-L _J-1＜0) time, adds 1.

By using the code table of relevant Δ L, the range information of each node of variable length code (Δ L _j).

The order of data array of distance and angle of will being used to encode is predefined for Δ L _j→ Δ θ _j→ Δ L _J+1→ Δ θ _J+1→ ....When the array of Δ L-Δ θ is configured to " 0.0_0.0_0.-2_+2.-2_0.+3_-5.0_0.0_0.+6 ", utilize Figure 26 (a) and code table length varying coded data string (b) as follows.

" 0.0_0.0_0.1011_1010.1011_0.11000_11101.0_0.0_0.11100 " → " 00000101110101011011000111010000111100 " (38 positions).

If distance component is represented to represent with the regular length of 10 positions with angle component with the regular length of 8 positions, so, need (position, 8 position+10) * position, 8 node=144, thereby, can be compressed to 26% to data volume by variable length code.

Figure 27 has shown the processing procedure of creating these code tables under off-line state.With reference to past transport information (step 90), at first, select the target road segmentation (step 91) of transport information.The position data that is included in the node in the target road segmentation converted to total curvature function representation (step 92) and calculate Δ L in each node of each segmentation according to the statistical value accounting equation _jWith Δ θ _j(step 93).Then, calculate Δ L _jWith Δ θ _jAppearance distribute (step 94) and according to Δ L _jThe code table of the relevant Δ L of appearance profile creation, and, Δ θ _jThe code table (step 95 and 96) of the relevant Δ θ of appearance profile creation.

In addition, Figure 28 has shown in order to send transport information, utilizes the processing procedure of the code table coding road shape data of creating.When receiving transport information (step 97), selection comprises the target road segmentation (step 98) of the occurrence positions of traffic events.The position data that is included in the node in the target road segmentation converted to total curvature function representation (step 99) and calculate Δ L in each node of each segmentation according to the statistical value accounting equation _jWith Δ θ _j(step 100).Then, by with reference to the code table data that are intended to the code table that the target road segmentation creates (or be intended to get near the link creation of the shape of target road segmentation code table), the Δ L in each node _jWith Δ θ _jConvert coded representation (step 101) to.Send the shape data (step 102) of the relevant target road segmentation of coding like this together with the data of relevant event location by the relative information representative in the target road segmentation.

Figure 29 and Figure 30 have shown the road shape data (Figure 29) and the event location data (Figure 30) that will send.The road shape data comprise code table data, the start node p1 of segmentation (node p1 and p2) that will be by coded representation) absolute coordinates, the absolute azimuth of node p1, distance L and (the Δ L that contains coding of the coded data node p1 and the p2 from node p1 to next node _jWith Δ θ _jBit string).

Receiving the take over party of data, by the identifying code table, helping curvature function by the data-switching of coded representation, thus with the treatment scheme of Figure 11 in identical mode reproduce the road shape data.Then, reproduce the map match of the road shape of shape and self numerical map, intended target roadway segment and according to the event location data is specified the traffic events occurrence positions in the target road segmentation.

Therefore, in method, not the coordinate points of resampling, but the data of the angle and distance of the relevant specified coordinate point of variable length code based on present embodiment, thereby, can reduce the transmission data volume of road shape data.

(the 6th embodiment)

In the 6th embodiment, to the position of the coordinate points of on road, resampling, angle component is remained unchanged, thereby the method for the distance component of only encoding is described.

As reference Figure 41 is described, by with respect to adjacent coordinates point (P _J-1) distance and this two dimension of angle can specify the coordinate points (P that is arranged on the road uniquely _J).In the 1st to the 4th embodiment, the position of the coordinate points of resampling remains unchanged the distance in this two dimension, thereby only encodes angle, to reduce the data volume that will send.In contrast, in the 6th embodiment, the position of the coordinate points of resampling remains unchanged angle, thereby only encodes distance, to reduce the data volume that will send.

Figure 31 shown make angle information fix (angle of deviation θ=constant) and the coding range information situation under the coordinate points of resampling.The processing of shape data of resampling is carried out as follows.

(1) from start node P ₀To terminal node road shape is followed the tracks of, and, next node P ₁Be arranged to angle of deviation reach predetermined angle theta (or-θ) position.

(2) when in (1), follow the tracks of and angle of deviation reach predetermined angle theta (or-θ) before with respect to start node P ₀Distance reach preset distance L _MaxThe time, next node P ₁Be arranged on that position.

(3) by a node P who in (1) or (2), determines ₁Be arranged to beginning edge, use the rule of (1) and (2) and determine next node P ₂Repeat this processing successively, to determine P ₃..., P _J....

Each the node P that resamples _JIn range information be used as adjacent node P _J+1Distance L _jWith distance statistics predicted value T _jBetween the prediction value difference Δ L of difference _j(=L _j-T _j) expression and Δ L _jWith, for example, 10m is that unit (can use such as with 50m or 100m other resolution as unit) is quantized.At this moment, distance statistics predicted value T _jBe defined by, for example, T _j=L _J-1Or T _j=(L _J-1+ L _J-2)/2.

Then, according to quantizing Δ L _jOccurrence frequency create the code table of relevant Δ L.At this moment, can calculate Δ L _jContinuous distribution, merged the code table of run length coding, RLC with establishment.

Figure 32 has shown the example of code table of the relevant Δ L of establishment like this.In this code table, regulation (if θ is positive, is a position of the sign of representing angle of deviation θ 0 when Δ L=0, if bear with θ, be 1) add in the code as additional bit, and regulation is when being provided with Δ L ≠ 0, a position containing the sign of representing angle of deviation θ and a position of sign of representing Δ L (if Δ L is positive, if be 0 and Δ L bear, be 1) two positions add in the code as additional bit.Like this, in definition of T _j=L _J-1Situation under,

When Δ L ≠ 0 is set,

If L _jGreater than L _J-1(L _j-L _J-1＞0), add 0, as additional bit of the sign of representing Δ L and

If L _jLess than L _J-1(L _j-L _J-1＜0) time, add 1, as an additional bit of the sign of representing Δ L, and,

If at P _J-2→ P _J-1Azimuthal left side (left curve) form P _J-1→ P _j, add 0, as additional bit of the sign of representing θ and

If at P _J-2→ P _J-1Azimuthal right side (right curve) form P _J-1→ P _j, add 1, as an additional bit of the sign of representing θ.

In the 4th embodiment, in the coordinate points of resampling, make under the situation that distance component remains unchanged, the example that changes distance component (section length of resampling) with segmentation is described.In addition, resampling, making under the situation that angle component remains unchanged, also can change the value of θ with segmentation.In this case, can with the 4th embodiment in identical mode, private code is used for shape data through code conversion, discern the value of θ in each segmentation.

Figure 33 has shown the processing procedure of creating code table under off-line state.In addition, Figure 34 has shown the code table coding road shape data of utilizing so establishment and the processing procedure that sends transport information.These processes are with the difference of the process of describing in the 3rd embodiment (Fig. 7 and Fig. 8), with fixed angle θ (or-θ) replace the resample shape data (step 112 of relevant target road segmentation of regular length L, step 121), the Δ L of each node that calculating is resampled, replace the Δ θ (step 114 of calculating each node of resampling, step 123), code table according to the relevant Δ L of the profile creation of Δ L, replacement is according to the code table (step 115 of the relevant Δ θ of the profile creation of Δ θ, step 117), other process is all identical.

In addition, Figure 35 has shown the road shape data that will send.These road shape data are that with the difference of the road shape data of describing they comprise the information of relevant sampling angle θ in the 3rd embodiment (Fig. 9), replace the information of relevant sampling section length L, and, comprise and contain the Δ L that is encoded into coded data _jBit string, replace the Δ θ contain coding _jBit string, others are all identical.

Receiving the take over party of data, by the identifying code table, helping curvature function by the data-switching of coded representation, thus with the treatment scheme of Figure 11 in identical mode reproduce the road shape data.Then, reproduce the map match of the road shape of shape and self numerical map, intended target roadway segment and according to the event location data is specified the traffic events occurrence positions in the target road segmentation.

Therefore, in the method based on present embodiment, the position of the coordinate points of resampling on road remains unchanged angle component, and a variable length code distance component is so that can reduce the transmission data volume of road shape data.

(the 7th embodiment)

In coding method, can select based on the expression of angle of deviation or based on the expression of prediction value difference, as the method for expression angle information, so that the shape of road converted to the shape data that has statistic bias based on the 7th embodiment.

As reference Figure 42 is described, based on angle of deviation θ _jExpression (Figure 42 (b) and (b ')) and based on angle of deviation θ _jPrediction value difference Δ θ _jThe situation of any angle information that is applied to coordinate points of expression under, can become exist the data of statistic bias to the road shape data-switching.

If statistic bias is bigger, the effect that reduces data volume according to variable length code is improved.Will be by angle of deviation θ _jThe angle information of denotation coordination point with by angle of deviation θ _jPrediction value difference Δ θ _jRepresent that same angle information compares, in general, latter event causes bigger statistic bias.

But, as shown in figure 36,, in a bit of, use the prediction value difference Δ θ of angle of deviation if in the road 40 that contains the straight line that between curve, forms _jThe expression angle information so, is provided with 0 ..., 0, θ ₁,-θ ₁, 0 ..., θ ₂,-θ ₂, 0 ... and if represent angle information with angle of deviation θ, so, be provided with 0 ..., 0, θ ₁, 0,0 ..., 0, θ ₂, 0,0 ....At the angle information of coordinate points angle of deviation θ _jUnder the situation about representing, with the angle information of coordinate points prediction value difference Δ θ _jSituation about representing is compared, and can have bigger statistic bias.

Therefore, best in some cases, should use angle of deviation θ _jThe coordinate points angle information that expression is resampled with constant distance L is to adapt to the variable length code of the shape that depends on road.

In method based on present embodiment, to represent data size that the shape of road obtains and represent that with the prediction value difference Δ θ of angle of deviation the data size of the shape acquisition of road compares with angle of deviation θ in order to carry out variable length code, with the less coded data of transmission data size in order to carry out variable length code.

At first, establishment is used in order to carry out variable length code angle of deviation θ _jThe expression road shape angle of deviation θ code table and be used in order to carry out variable length code angle of deviation θ _jPrediction value difference Δ θ _jThe Δ θ code table of the shape of expression road.

Figure 37 has shown that process and Figure 38 of creating angle of deviation θ code table have shown the process of creating Δ θ code table.Process shown in Figure 38 is identical with process (Fig. 7) among the 3rd embodiment.In addition, the difference of process shown in Figure 37 and process shown in Figure 38 is that angle of deviation θ is used to replace Δ θ.

Figure 39 has shown these code tables coding road shape data that utilization is created and the processing procedure that sends transport information under off-line state.

Step 130: receiving traffic information.

Step 131: selection comprises the target road segmentation of the occurrence positions of traffic events.

Step 132: with regular length L resample the target road segmentation the road shape data and node is set.

Step 133: the position data that node is set is converted to the total curvature function representation.

Step 134: then, by create code data and the computational data size (A) of relevant θ with reference to the code table of relevant θ.

Step 135: then, by create code data and the computational data size (B) of relevant Δ θ with reference to the code table of relevant Δ θ.

Step 136: data size (A) is compared with data size (B), the value of " coded data " during " angle is represented mark and label " that the angle of using less data size is represented and indication application point of view like this is represented and application point of view are represented is arranged to the shape data that will send.

Step 137: send the shape data of relevant coding target road segmentation together with the data of the relevant event location of in the relative information of target road segmentation, representing.

Figure 40 has shown the road shape data that will send.The road shape data comprise that " angle is represented mark and label " that relevant indication application point of view represents is (when the expression of using based on angle of deviation θ, be 0, with when using when representing based on predicted value difference Δ θ, be 1) information and the information of " coded data " during relevant application point of view is represented.

Receiving the take over party of data, with reference to based on code table by " angle is represented mark and label " specified message, reconstruct θ or Δ θ from code represented " coded data ", thus the position data of each node is converted to the total curvature function.Processing subsequently is identical with among the 3rd embodiment those, reproduce the road shape data, reproduce the map match of shape and self numerical map,, in the target road segmentation, specify the traffic events occurrence positions with the intended target roadway segment with according to the event location data.

Therefore, in method, based on the expression of angle of deviation with based on any method that is selected as the expression angle information of the expression of prediction value difference based on present embodiment.In view of the above, can further reduce the data volume that will send.

Also can be applied to the map datum main body based on coding method of the present invention.In addition, this method also can be applied to map datum and transmit service in transfer on the Internet (for example, utilizing the client-server type map display system of map vector) or map datum.

In addition, to send to the running orbit data under the situation at center, also can utilize the coding method packed data at the Hot Report that sends for airborne machine or flowing car data (FCD) from the vehicles.

In addition, will also using, utilize code table that data compression is carried out in the expression of node strings with under batten compression method compression shape vector and its situation of data transmission as relevant each node strings based on coding method of the present invention.

And, can also be applied to send the situation of the shape data in the zone (polygon) on the numerical map based on coding method of the present invention.For example,, send under the situation of the weather forecast in this zone, send the shape data of the boundary line of polygonal shape, so that the take over party can specify polygon specifying a polygon.Under the situation of the shape data that will send the boundary line, can use based on coding method compression of the present invention and send data volume.At this moment, there is no need accurately to specify polygonal shape in the zone of using weather forecast, the take over party can omit the matching treatment with the shape of numerical map.

Shown code table is an example, and may not be always best.In fact, must check variable (θ _j, Δ θ _j, L _j) distribution, thereby utilize the Hofman tree structure to create code table.

For coding techniques, there are various methods, for example, fixed length character compression method, run length method, Shannon-Fan Nuo coding method, huffman coding method, self-adaptation type huffman coding method, arithmetic coding method and dictionary method (LHA method).In this manual, also can use these coding methods.Though the situation to generating code table under off-line state is described,, also can utilize self-adaptation type huffman coding method or arithmetic coding method, under presence, encode.

Though only specifically described some embodiment of the present invention here,, apparent, can do many modifications to them, and without departing from the spirit and scope of the present invention.

The Japanese patent application that the present invention is based on Japanese patent application 2001-134318 number that proposes May 1 calendar year 2001 and propose on July 19th, 2002 2001-220061 number is quoted, for your guidance hereby.

Industrial Applicability A

From top description, can obviously find out, in based on coding method of the present invention, can be effectively Be compressed in the data volume of the shape vector on the numerical map. Therefore, sending out based on positional information of the present invention In delivery method and the equipment, in the situation of the shape vector that will send numerical map, can subtract considerably Send less data volume. The recipient, reconstruct shape data and carry out map match from receive data, with Just can accurately specify the shape vector of transmission like this.

Claims (8)

1. the shape data coding method of data of the shape of geological information is represented in a coded representation, and the method comprising the steps of:

Utilize node strings to represent described shape, wherein, each node in the described node strings is by having at least two positions of elements information representations;

Conversion about an element in the described element of the described positional information of each node of described node strings providing, so that describedly concentrate about predetermined value greater than other elements in the described element of described positional information through data converted through data converted; And

It is described through data converted to encode.

2. shape data coding method according to claim 1, wherein, use about information representation about described other elements in the described element of the described positional information of described node, and use about information representation from the angle of the extended straight line of adjacent node about a described element of the described element of the described positional information of described node from the distance of adjacent node.

3. shape data coding method according to claim 2,

Wherein, represent with angle of deviation about the information of angle, and this angle of deviation of encoding.

4. shape data coding method according to claim 2,

Wherein, represent with the prediction angle difference between the predicted value of this angle and this angle about the information of angle, and this described prediction angle difference of coding.

5. shape data coding method according to claim 2,

Wherein, use about the difference of the predicted value of this distance about the described information of this distance and to represent, and this difference of encoding.

6. shape data coding method according to claim 1 and 2,

Wherein, represent described node again, so that in the predetermined segment of described shape, get fixed value about described other elements in the described element of the described positional information of described node.

7. shape data coding method according to claim 6,

Wherein, represent node again with such position: the distance from adjacent node is constant, and only represents with angle information about the described positional information of described node.

8. shape data coding method according to claim 6,

Wherein, represent node again: get not varied angle from the angle of deviation of the extended straight line of adjacent node, and only represent with range information about the positional information of described node with such position.

Images