JPH037488A - White map forming method in catv network design - Google Patents

Abstract

PURPOSE:To generate a white map from a map on market by converting picture information obtained from a house map or the like into graph information, discriminating length of line of line segments forming each graph in the graph information so as to separate map major information, characters and symbols. CONSTITUTION:Line segments of all constitution lines are discriminated based on a prescribed value of l6<Lo<l1 to separate a graph whose constituting lines are smaller than Lo from a graph whose length of one side constituting one polygonal line is larger than Lo. That is, since a graph of a house (constitution lines 141-144) is larger than Lo, it is discriminated that the line is not a character nor a symbol, and since the character of the name of the household master of the house (constitution lines 151-152) is smaller than Lo, it is discriminated as a character or a symbol. Similarly, the graph is discriminated based on a threshold level L1 (l1<L1), the contour line 14 of a house and the border of a road constituted with longer lines are separated. Thus, a white map is formed from a house map on market.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an electronic blank map creation method, and more particularly to a blank map creation method in CATV network design.

[Prior art] One of the conventional CATV network design methods is scale 11.
On a blank map showing roads and houses of 500 to 1/1000 scale, coaxial cables, amplifiers,
Transmission equipment such as branches and distributors are placed while calculating their characteristics. This design method has the disadvantage of requiring a huge amount of work since the number of residential units that make up the system is usually several thousand or more, and furthermore, it requires a huge amount of work as it requires the creation of related drawings such as level diagrams of transmission signals. It has the disadvantage of requiring

Therefore, in recent years, cases in which this network design is performed using CAD are increasing. The methods of inputting the blank map, which is the basis for this CATV network design, into the CAD system are: (1) using a digitizer to input information on roads, houses, etc. sequentially; (2) using a CD ROM to input information on roads, houses, etc. However, the first method (1) has the drawback of requiring a huge amount of effort, and the second method Since commercially available databases are general-purpose, they have a rich amount of information, and when used for CATV network design, they have drawbacks that result in increased costs.Furthermore, the database for the required region is not available. These blank map creation work conditions are hindering the use of CAD for CATV network design.

[Problems to be solved by the invention] This invention has been made in view of the above points, and
The purpose of the present invention is to provide a method for creating high-quality electronic blank maps from commercially available maps, which is realized by a small-scale system.

[Means and effects for solving the problem] The present invention converts image information obtained from a residential map or the like into graphic information, and calculates the line lengths of line segments constituting each figure in this graphic information,
Alternatively, map main information, characters, symbols, etc. can be obtained by discriminating quantities derived secondarily from the line lengths of line segments, such as the total line length, the line length ratio of vertical and horizontal lines, the area of a closed loop formed by multiple line segments, etc. It is characterized by separating.

The above configuration enables the use of a commercially available residential map or the like as an image information source, and also functions to process image analysis and recognition of image information obtained from the residential map or the like relatively easily and at high speed.

[Embodiment] First, an example of a system suitable for carrying out the present invention will be described with reference to FIG. In the following description, a two-dimensional pattern containing only visual density information will be referred to as an image, and a two-dimensional pattern consisting of points or lines having coordinate values will be referred to as a figure. Therefore, the data format of image information is rask information, and the data format of graphic information is vector information.

In FIG. 1, reference numeral 1 is an image scanner that reads images from residential maps or city planning maps, etc., 2 is a computer that analyzes and recognizes the image information read by the image scanner 1, and converts it into graphic information; 3 is the computer that reads the image information from the image scanner 1; This is a display device that displays image information or graphic information as appropriate, and this display device 3 includes a display section 4 and a keyboard section 5.
It is designed to perform image analysis and recognition using an interactive method.

Next, an example of an original image input to the image scanner 1 will be described with reference to FIG. 2, which illustrates a residential map.

In the figure, 6 is the outline of the house, 7 is the character indicating the name of the household head of the house, 8 is the road boundary, 9 is the boundary line of matters/prefectures, etc., and 10 is the road sign symbol indicating one-way traffic. ,1
1 is the character indicating the name of the road, 12 is a part of the address T name number, 13 is the house address, and the outline of the house 6,
Road boundaries 8 are information necessary for CATV network design, including characters 7 that indicate the name of the head of the household, boundaries 9 of boundaries such as matters/prefectures, road sign symbols 1 that indicate one-way traffic, and characters 11 that indicate the name of the invention. , the first part 12 of the address T name number and the house number 13 of the residential address are unnecessary information.

From Figure 2, it can be understood that images such as housing maps or city planning maps, including the characters and symbols written on them, are visually composed of lines with binary density. For example, a "point" on this residential map is recognized by the image scanner 1 as a set of a plurality of dots, and similarly, a '° line' on the residential map is recognized as a pattern.

Further, the density which should be binary is recognized as the density of multi-value gradation.

Since such image information has a large amount of information and is not suitable for computer processing, density threshold processing,
Thin line processing, vector conversion processing, etc. will be performed. However, whether or not to distinguish these processes as pre-processing is a matter of the algorithm, and a well-known method can be adopted, so a detailed explanation will be omitted.

FIG. 3 illustrates a part of the graphic information after the image information has been subjected to the above-described density threshold processing, MA line processing, vector conversion processing, etc., and numeral 14 is the outline of a house, 141 to 145 is its constituent line, 15 is a character indicating the name of the head of the household, and 151 to 152 are its character constituent lines. In order to clearly distinguish between the image information in FIG. 2 and the graphic information in this figure, different symbols are used for the outline of the house and the characters indicating the name of the head of the household in each figure.

Next, one embodiment of the present invention will be described with reference to FIG. 3 and the flowchart of FIG. 4 illustrating the image analysis/recognition method of the present invention.

In Fig. 3, the contour line 14 of the house is a single-stroke polygonal line starting from one corner and ending at that starting point, and similarly, the characters 15 indicating the name of the householder of the house are also composed of a similar polygonal line. It is understood that The configuration of each figure after being converted into this figure information t! 141-14
5,151 to 152 have coordinate values, so their lengths can be easily determined. Therefore, increase the line length of all component lines by 1° for each figure. By performing discrimination using a predetermined value Lo of 9-1 (expressed as line length separation processing on the flowchart in Figure 4), even if the length of one side constituting one polygonal line is one, It is possible to separate large figures from figures whose constituent lines are all smaller than Lo, that is, the figure of a house has its constituent lines 141 to 144.
is larger than LO, so it is determined that it is not a character or symbol figure, and the character shape of the name of the head of the household is determined to be a character or symbol figure because its character constituent lines 15+ to 152 are smaller than Lo, respectively. Separation (expressed as house recognition processing in flowchart 1 to above of FIG. 4) is performed.

Similarly, by setting a predetermined value such that l I < L 1 and performing discrimination using 1, it is possible to separate the residential contour line 14 from the road boundary line formed by a longer line. In addition, the emphasis on residential maps illustrated in Figure 2, dotted lines 9 of boundaries such as prefectural borders, road sign symbols 1 indicating one-way traffic, and name of the invention 1
1. Most of the unnecessary information such as the first part 12 of the address T name number and the street number 13 of the residential address can be separated. It should be noted that the discrimination value LOL+ may be fixed or may be made variable through an interactive method.

The graphics separated in this way are hierarchically stored and called up as appropriate, and can be used for CATV network design by being displayed independently or simultaneously.

Next, another line length separation method of the present invention will be explained.

In the straight line p=xcO8θ+ySINθ (where ρ is the length of the perpendicular line drawn from the origin and θ is its angle),
By parameter conversion, all points on this straight line (x, y
) are concentrated at one point on the ρ-θ plane.

In other words, the coordinates (Xl, yl) of the first point (starting point) a+ of a certain line segment a and the coordinates (X2
1 y2 ), when ρ and θ of the line segment a are determined, p3 = x3 CO3θ+y3 S ■ Nθ is derived using the coordinate value (X3y3) of the third point a3 and, for example, the previous θ, and the following Similarly, ρ for the nth point. =xnCO3
θ+ynSINθ are sequentially derived. Mathematically, it is understood that these ρ, ρ3 to ρ□ are all equal, and that different ρ can be obtained for points that are not on the line segment a.

However, the image information actually output by the image scanner 1 or the image information of a straight line after thin line processing is ρ obtained by the previous calculation due to the quantization error.

ρ3～ρ. contains some errors. Therefore, any △
With ρ being the tolerance, points within Δρ of the image information, which is a set of points output by the image scanner 1, must be considered as points constituting a specific line segment.

Detect the starting point using an arbitrary method, determine ρ and θ from the starting point and points that follow it, then calculate θ or ρ from the coordinate values of the continuous points and either ρ or θ, and calculate Δθ.
Alternatively, a straight line is recognized by determining whether it is within Δρ or not, and when the tolerance is exceeded, that point is determined to be the extreme point on the image. Since the starting point and the extreme point on the image have a one-to-one correspondence with the coordinates, the coordinate values of the starting point and the extreme point of the straight line are recognized and stored together with the line segment code. Note that the number of points within Δθ or 69 until the pole point is recognized is approximately proportional to the length of the line segment, so the line length can also be recognized at the same time.

Furthermore, when the above process is repeated for other line segments starting from this extreme point until there are no consecutive points, the line lengths of all the constituent line segments of each piece of image information are recognized and stored.

Thereafter, the maximum value of the constituent line segments of each piece of image information is discriminated using a plurality of threshold values, so that the graphic information is hierarchized in the same way as in the previous embodiment. That is, characters and symbols are recognized as figures with the shortest maximum line length of their constituent lines, and house figures are recognized as the figure with the second shortest maximum line length of their constituent lines.
A road boundary line is recognized as a figure with the longest maximum line length of its constituent lines.

Still another embodiment of the present invention will be described with reference to FIG.

In the figure, reference numerals 16 and 17 are road boundaries, and 18 is a residential outline. The contour line 18 of the house forms a single-stroke polygonal line that starts at one corner and returns to the starting point, that is, forms a closed loop, whereas one of the road boundary lines 17 does not form a closed loop. Due to this feature, it is possible to separate the figures forming a closed loop into house figures and the others as road figures. Note that a figure forming a closed loop, such as the boundary M16 of another road, can be separated depending on whether there is another figure within the closed loop.

In addition, even if there are no other shapes in the closed loop, for non-residential shapes such as vacant lots and station platforms, calculate the area of the closed loop and the sum of the line lengths of each side of the closed loop, or House figures and road figures or other figures can be separated by discrimination processing of the ratio of vertical and horizontal line lengths.

[Effects of the Invention] As described above, according to the present invention, in addition to the effects of automation, (1) image information can be obtained from commercially available residential maps.

(2) Since line figures are handled, the amount of data required to be processed and the amount of stored data is small.

(3) It is easy to hierarchize analysis data.

(4) Suitable for an interactive method in which setting values are changed while referring to the converted figure.

It is possible to provide a method for creating a blank map in CATV network design that has the following remarkable effects.

[Brief explanation of the drawing]

Fig. 1 is a diagram illustrating an example of a system suitable for carrying out the present invention, Fig. 2 is a diagram illustrating a commercially available residential map that is an original image, and Fig. 3 is a diagram illustrating a residential map converted into a line diagram. FIG. 4 is a flowchart of the embodiment of the invention; FIG. 5 is a partially enlarged view of a residential map converted into a line figure; It is a figure explaining other Example of this invention. 1... Image scanner 2... Computer body 3... ......Display device 4...Display section 5... ... Keyboard part 6.14.18 ... Outline of the house 7.1
5......Character 8 indicating the name of the head of the household
．． 16.17 Road boundary line 9
・・・・・・・・・・・・ Matters/Boundaries such as prefectural borders 10・
・・・・・・・・・・・・Road sign symbol 11 indicating one-way street ・・・・・・・・・・・・・Character 12 indicating the name of the road・・・・・・・・・Address T Name number 1 part 13・・・・・・・・・・・・Residence address number 141-145・・・・・・House outline composition lines 15, ~152 ...Character composition lines

Claims (2)

[Claims] (1) Image information obtained from residential maps, etc. is converted into graphic information, and main map information and characters, symbols, etc. are separated by line length discrimination of the line segments that make up each figure in this graphic information. A blank map creation method for CATV network design. (2) Whether or not the line segments that make up each figure in the figure information form a closed loop, whether or not other figures exist within the closed loop, and the line lengths of the line segments that make up each figure. It is characterized by separating main map information from characters, symbols, etc. by discrimination of quantities derived secondarily from, for example, the total line length, the line length ratio of vertical and horizontal lines, and the area of a closed loop formed by multiple line segments. The blank map creation method for CATV network design according to claim 1.