JPH10269354A - Method for eliminating noise pixel, geographic image processor and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a geographic image processor which easily detects a noise pixel that is always included in satellite image data and eliminates it. SOLUTION: An image range specifying part 31 specifies an objective range on a geographic image that is generated based on satellite image data. A pixel eliminating part 34 eliminates a color pixel that satisfies a condition which regards it as a noise pixel that is set by a condition setting part 33 from a geographic image within a specified objective range, that is, the condition on which the number of pixel connections in a setting direction (four-connection direction/eight-connection direction) is below a connection upper bound value from the area.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a geographic image processing technique which is the basis of map data, and more particularly, to a plurality of generations of satellite image data, that is, a plurality of satellite image data obtained by photographing the same point. The present invention relates to a technique for generating a geographic image from each of them, removing noise pixel components included in each geographic image, and improving the accuracy of extracting difference information between the geographic images.

[0002]

2. Description of the Related Art Map data representing geographical information such as residential land, farmland, roads, rivers and the like, and arrangement information of buildings on the surface of the ground, is converted into a database, and necessary map data is appropriately displayed on a display unit as an image or printed. For example, a service system for providing a consumer is known. In such a service system, map data is generated by periodically acquiring aerial photographs, satellite photographs (hereinafter, aerial photographs, etc.), or survey data on the ground surface, and the contents of the map data are changed over time. If it has changed, the operator has searched and corrected the corresponding map data from the database. In recent years, instead of paper-based map data, pattern recognition of aerial photographs etc. using a computer has been performed, and the generation of individual map data and the detection of changes in the contents of multiple map data have been automated. Attempts have been made to improve the correction accuracy of map data.

[0003]

In the case where map data is generated by performing pattern recognition on aerial photographs and the like, accuracy of pattern recognition and recognition processing time become problems. Also, the presence of noise present in aerial photographs and the like cannot be ignored. That is, an aerial photograph or the like is photographed including a moving object such as a human, an animal, a vehicle, a train, and the like, in addition to a terrain and a building that are originally intended to be photographed. Depending on weather conditions, clouds, rain, and the like may be photographed. The existence of these moving objects and clouds is harmful noise from the viewpoint of detecting a change in the content of map data, and even when capturing the same ground surface that has not changed between multiple generations, It is recognized as having changed due to the presence of the body or the like.

[0004] It is an object of the present invention to provide a computer-based noise elimination method that easily detects noise always included in a geographic image and increases the accuracy as map data. Another object of the present invention is to provide a geographic image processing apparatus capable of accurately detecting a change in the contents of map data and easily grasping a changed portion on a geographic image, and a general-purpose computer using the geographic image processing apparatus. An object of the present invention is to provide a recording medium for realizing the same.

[0005]

The noise elimination method of the present invention which solves the above-mentioned problems includes an image of the ground surface photographed by an artificial satellite, condition information at the time of photographing, weather information at the time of photographing, etc. Pixels of the geographic image generated based on the image information are referred to as satellite image data. The pixels constituting the geographic image are binarized as necessary, and the number of connected pixels in the designated direction is equal to or less than a predetermined upper limit of connection. Is detected, the detected colored pixel is regarded as a noise pixel and is removed from the corresponding area. Such processing is executed on the computer.

[0006] Since satellite image data contains high-density image information, the use of this makes it easier to create map data than in the conventional case of performing aerial image processing and the like. . In addition, a geographic image generated based on satellite image data is composed of a set of closed area elements or line elements, and when the geographic image is enlarged and viewed at a pixel level, a closed area element such as a building or a road edge is obtained. While the connection direction of line elements such as is regular and the number of connections is relatively large,
The connection direction of noise pixels is irregular, and the number of connection of noise pixels is relatively small. Focusing on this point, the method of the present invention detects noise pixels from the geographic image and removes them.

[0007] In order to solve the above-mentioned other problems, the present invention removes noise pixels and removes contents between a plurality of geographic images (geographic images based on satellite image data of a plurality of generations, the same applies hereinafter) at the same point. There is also provided a geographic image processing device with improved change detection accuracy. The geographic image processing apparatus sets an image range specifying unit that specifies a target range on a geographic image generated based on the satellite image data, and sets a connection direction and a connection upper limit value of colored pixels to be regarded as noise pixels. A condition setting means for detecting, from the geographic image within the specified target range, a color pixel whose number of pixels connected in the set direction is equal to or less than the connection upper limit value, and removing the detected color pixel from the corresponding area; And a removing unit.

[0008] If necessary, difference information between a plurality of geographic images representing the same point from which the noise pixels have been removed by the pixel removal means is generated and visualized, preferably, geographical information within the target area is generated. There is also provided difference information visualization means configured to be highlighted on the image. Furthermore, the image processing apparatus further includes an element determining unit that determines whether the geographic image in the target range is a closed region element or a line element, and the condition setting unit determines that the first direction has priority when the geographic image in the target range is a closed region element. In the case of a line element, the second direction is preferentially set.

A recording medium according to the present invention for solving the above-mentioned other problems is a computer-readable recording medium in which a program for causing a computer to execute the following processing is recorded. (1) a process of setting a connection direction and a connection upper limit value of a colored pixel to be regarded as a noise pixel; (2) pixel connection in the set direction from a geographic image within a target range specified by the image range specification unit A process of detecting a color pixel whose number is equal to or less than the connection upper limit value;

[0010]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a configuration diagram showing an embodiment of the geographic image processing device of the present invention. The geographic image processing apparatus 1 includes an image processing unit 2, which is realized on the information processing apparatus by an information processing apparatus such as a general-purpose workstation or a personal computer reading a computer program and executing the computer program. And a noise removal processing unit 3. This computer program is usually incorporated in an internal storage device or an external storage device of the information processing apparatus, and is distributed integrally with the information processing apparatus. Recording medium, for example, a CD-ROM (compact disk type RO)
M) or an FD (flexible disk), or may be distributed through a machine-readable communication network or the like, and may be installed in the information processing apparatus at the time of system construction and executed at any time.

The image processing unit 2 is connected with a data input device 4 composed of a pointing device and a keyboard and a storage device 5 composed of a large-capacity hard disk and the like, and a CRT for confirming condition inputs and processing results. Are connected.

The storage device 5 stores satellite image data (raw data) in addition to the storage area of the computer program.
Storage area, a storage area for geographic image data (data on which a geographic image is based; hereinafter, a geographic image may include this geographic image data in some cases), noise pixels, and necessary image processing. A storage area for the application data is formed. The application data is stored in the geographic image processing apparatus 1 and other image processing systems.
This data has been processed so that it can be used after the fact.

As shown in FIG. 2, the image processing unit 2 includes a data conversion processing unit 21, a data input / output processing unit 22, a difference analysis processing unit 23, an image enhancement processing unit 24, a display control unit 25, and each of these units. And a main control unit 26 that controls the entire system.

The data conversion processing section 21 has a storage device 5
The satellite image data (raw data) stored in or stored at any time from a provider is taken in, converted into a data format that can be processed by the geographic image processing apparatus 1, and binarized for convenience of post-processing. To generate the geographic image data. The generated geographic image data is represented as a set of pixels. There are as many pixels as the number of x and y components in two-dimensional coordinates. In the case of a binarized geographic image, a pixel having a pixel value of “1” including its edges in a closed region, that is, a colored pixel Are continuous, and pixels having a pixel value of “0” outside the closed region (that is, no colored pixels exist) are continuous. Here, the pixel value is an identifier given to each pixel by the data conversion processing unit 21.
“0” is black and “1” is white. Binarization allows the image to be represented by a connected distribution of these pixel values, thus speeding up the subsequent processing. However, it is a matter of course that the present invention is not limited to only binarized images, and that the present invention can be realized by using a plurality of colored pixels as they are.

[0015] The generated geographic image data is stored in a storage device 5.
And is also sent to the noise removal processing unit 3. The data input / output processing unit 22 includes the satellite image data (raw data) between the noise removal processing unit 3 and the storage device 5,
The input and output of the geographic image data and the application data generated by the image processing unit 2 and the transfer of the condition data and the like input by the operator to the data input device 4 to the noise removal processing unit 3 are performed.

The difference analysis processing unit 23 includes a plurality of generational geographic images of the same point from which noise pixels have been removed by the noise removal processing unit 3 (a plurality of geographic images of the same point taken at different times and different time zones). The difference between the images is detected by analyzing the distribution state of the pixel values between (1) and (2). The image enhancement processing unit 24 highlights the analysis result of the difference analysis processing unit 23 on the display of the display unit 6. is there. The display control unit 25 performs control for that. Examples of the mode of the highlight display include, for example, changing the color of the change on the geographic image, displaying the change, and blinking the display. The data to be highlighted is stored in the storage device 5 as application data. In addition,
The application data may be processing result data of the difference analysis processing unit 23.

The display control unit 25 performs not only display control for the above-mentioned highlight display on the display unit 6 but also display control for window display for prompting an operator to input condition data described later.

As shown in FIG. 3, the noise removal processing section 3 includes an image range specifying section 31, an element type determining section 32, a condition setting section 33, and a pixel deleting section 34.

The image range specifying unit 31 is, for example, a range of a geographic image for performing a noise pixel removal process specified by a data input device (pointing device) 4 from a geographic image displayed on the display unit 6. Is specified. This can be realized by a bitmap coordinate calculation process corresponding to the display. In addition,
The image range specifying unit 31 may be prepared as a function of the image processing unit 2.

The condition setting section 33 sets a condition value for searching for a noise pixel on a geographic image. Specifically, for example, a dialog window for selecting “4-way connection” or “8-way connection” and an input window for a connection upper limit value (natural number) are displayed on the display unit 6, and the operator Recognizes the input data. Then, the recognized data is set as a condition value at the time of search. here,
“4-way connection” means that a pixel having a pixel value “1” is a pixel in FIG.
(A) refers to a relationship of connection in four directions indicated by arrows,
The “direction connection” means that the pixel having the pixel value “1” is the pixel in FIG.
Are connected in eight directions indicated by arrows. The “connection upper limit value” refers to the maximum value of whether or not the number of pixels connected in these directions is less than or equal to a noise pixel.

The effect of setting the connection direction is that the element type of the geographic image, that is, the image to be captured is a closed area element such as a building, a farmland, a vacant lot, or a line element such as a road edge or a river edge. It depends on what you have.

For example, in the case of a closed area element as exemplified in FIG.
There are one two-connected figure (regular image) and two one-connected figures (noise pixels N). If the condition of “8-way connection” is set, the fourteen-connected figures (noise pixels N are mixed) Image) becomes one. In other words, the noise pixel N cannot be removed from the closed area element by the “8-way connection”, but the noise pixel N is removed by the “4-way connection”, and pixels within the range of the thick line are extracted as a normal figure. be able to. In the illustrated example, the “connection upper limit value” when the noise pixel N is completely removed is “1”.

On the other hand, in the case of the line element exemplified in FIG. 5B, when the condition of "four-way connection" is set, three two-connected figures (one is due to the noise pixel N. One of the three connected figures is unspecified), and when the condition of "8-way connection" is set, one figure of seven connection and one figure of two connection (noise pixel N) are set. Become.
In other words, the noise pixel N cannot be specified and removed from the line element in “4-way connection”, but the noise pixel N can be specified in “8-way connection”. In the illustrated example, the “connection upper limit value” when completely removing the noise pixel N is any one of “2” to “6”.

As described above, the element type determination section 32 determines whether the geographic image within the target range is a closed area element or a line element. As a result of the determination of the element type, the condition setting unit 33 preferentially displays “4-way connection” in a window when the element is a closed area element and preferentially displays “8-way connection” in a window when the element is a line element. , The condition value can be set according to the element type. FIG. 6 shows this state. FIG. 6A shows a case where a closed area element exists in the target range.
(B) is an example in the case where a line element exists in the target range.
In the case of this example, a condition input window as shown in the lower part of the figure is displayed on the display of the display unit 6.

The pixel deletion unit 34 analyzes the geographic image within the target range at the pixel level, and searches for a connected pattern (pixel group) that matches the set condition value.
If there is a matching connection pattern, the corresponding pixel is regarded as a noise pixel and is deleted from the corresponding area.
After that, the deletion result is stored in the difference analysis processing unit 23 of the image processing unit 2.
And the display is displayed on the display unit 6 so that the operator can confirm the degree of deletion.

Next, a specific operation of the geographic image processing apparatus will be described with reference to FIG. FIG. 7 mainly shows a processing procedure of the noise removal processing unit 3.

First, the geographic image is displayed on the display unit 6 (step S101). Noise removal processing unit 3
Detects that the operator who has viewed the display display has designated (input a condition) the target geographic image range using the pointing device (step S102: Ye).
s), a target range on the geographic image, that is, a processing region is specified by the specified data, and an element type in the region is determined (step S103). Thereafter, as shown in the lower part of FIG. 6A or 6B, a dialog window for prompting the user to input the conditions of the connection direction and the upper limit of the connection is displayed (step S104), and it is detected that the condition data has been input. Sets a condition value, searches for a connected pattern that meets the condition value, and deletes the corresponding pixel (step S105: Yes, step 106).

The image processing unit 2 displays the result of pixel deletion on the display unit 6 as an image. At this time, if the operator indicates that the result is not good, that is, if the degree of pixel deletion is not sufficient, or In the case of excessive deletion, the noise removal processing unit 3 returns to the processing of step S104 and prompts re-input of the condition data (step S107: No). on the other hand,
If the pixel deletion result is good (step S107: Ye
s) The extracted geographic image is extracted and output to the image processing unit 2 (step S108). If there is a next target range, the process returns to step S102, and the same process is executed by changing the target range (step S109: Yes). If there is no more target range, the process ends (step S
109: No). The geographic image from which noise pixels have been removed in this manner is subjected to difference detection between the geographic images in the image processing unit 2.

[0029]

As is apparent from the above description, in the noise removal method of the present invention, a geographic image is generated using satellite image data, and noise pixels always included in the geographic image are connected in the pixel connection direction and the connection upper limit. A search is performed based on the value, and when there is a pixel regarded as a noise pixel, it is deleted, so that there is an effect that a geographic image that can be used as high-accuracy map data can be easily and quickly generated. .

Further, in the geographic image processing apparatus of the present invention, since the difference between the geographic images from which noise pixels have been removed is detected and highlighted, the operator can easily grasp a changed portion on the geographic image. As a result, there is an effect that labor saving of database maintenance work is more thoroughly performed.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a geographic image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram of an image processing unit according to the embodiment.

FIG. 3 is a block diagram of a noise removal processing unit according to the embodiment;

4A is an explanatory diagram of four-way connection, and FIG. 4B is an explanatory diagram of eight-way connection.

5A is an explanatory diagram illustrating an example of a closed region element, and FIG. 5B is an explanatory diagram illustrating an example of a line element.

FIG. 6A is an example of a condition data input window when the element type of a geographic image in a target range is a closed area element;
(B) is an explanatory view showing an example of a condition data window in the case of a line element.

FIG. 7 is a flowchart illustrating an operation procedure of the geographic image processing apparatus according to the embodiment.

[Explanation of symbols]

 Reference Signs List 1 geographic image processing device 2 image processing unit 3 noise removal processing unit 4 data input device 5 storage device 6 display unit 21 data conversion processing unit 22 data input / output processing unit 23 difference analysis processing unit 24 image enhancement processing unit 25 display control unit 26 Main control unit 31 Image range specifying unit 32 Element type determining unit 33 Condition setting unit 34 Pixel deleting unit

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Susumu Fujikawa 6-81 Onoecho, Naka-ku, Yokohama-shi, Kanagawa Prefecture Hitachi Software Engineering Co., Ltd.

Claims (8)

[Claims] 1. A color pixel whose number of connected pixels in a designated direction is equal to or smaller than a predetermined upper limit of connection is detected from a geographic image generated based on satellite image data, and the detected colored pixel is regarded as a noise pixel and a corresponding area is detected. A method for removing noise pixels, wherein a process of removing noise pixels is executed on a computer. 2. A process executed on the computer,
2. The method according to claim 1, further comprising a step of binarizing constituent pixels of the geographic image before detecting the colored pixels. 3. An image range specifying means for specifying a target range on a geographic image generated based on satellite image data, and a condition setting means for setting a connection direction and a connection upper limit value of colored pixels to be regarded as noise pixels. And detecting, from the geographic image within the target range specified by the image range specifying unit, a color pixel in which the number of connected pixels in the set direction is equal to or less than the connection upper limit value, and regards the detected color pixel as a noise pixel. A geographic image processing apparatus, comprising: 4. The apparatus according to claim 1, further comprising: difference information visualization means for generating difference information between a plurality of geographic images representing the same point from which the noise pixels have been removed by the pixel removal means, and visualizing the generated difference information. The geographic image processing apparatus according to claim 3, wherein 5. The geographic image processing apparatus according to claim 4, wherein the difference information visualizing means is configured to highlight the generated difference information on a geographic image within the target range. . 6. The condition setting means, when the pixels are rectangular pixels arranged in a matrix, include a first direction representing four sides of each pixel or a second direction representing sides and angles.
The geographic image processing apparatus according to claim 3, wherein one of the directions is selectively set. 7. The apparatus according to claim 1, further comprising: an element determining unit that determines whether the geographic image within the target range is a closed area element or a line element, wherein the condition determining unit determines a result of the determination by the element determining unit;
7. The geographic image processing apparatus according to claim 6, wherein the first direction is preferentially set for a closed area element, and the second direction is preferentially set for a line element. . 8. A process for specifying a target range on a geographic image generated based on satellite image data; a process for setting a connection direction and a connection upper limit value of colored pixels to be regarded as noise pixels; Processing for detecting a color pixel whose number of connected pixels in the set direction is equal to or less than the connection upper limit value from the geographic image within the target range specified by the above, and regards the detected colored pixel as a noise pixel and removes it from the corresponding area A computer-readable recording medium on which a program for causing a computer to execute a process is recorded.