KR102041320B1 - Precision-Location Based Optimized 3D Map Delivery System - Google Patents

Abstract

According to the present invention, a precision location-based optimized 3D map delivery system comprises: a data input module configured to receive an actual image necessary for making a map and multiple pieces of map coordinate information stored by image from a data server; an image optimization module including an image analyzing unit configured to reconstruct the actual image into an intermediate image by filtering and analyzing the actual image, and an optimizing unit configured to calculate a detail level value of the intermediate image and perform allocation and rendering of polygons included in the intermediate image depending on the detail level value to generate a 3D image, and store the 3D image in the data server; a road position correction module including a coordinate display unit configured to display the multiple pieces of map coordinate information displayed as points on a road in the 3D image, and a position aligning unit configured to align lane positions of the road in the 3D image in accordance with a reference line which is one among multiple lines connecting the points in an extension direction of the road; and a map delivery module configured to combine background images based on the 3D image, in which the lane positions are aligned, to generate and output a 3D map.

Description

Precision-Location Based Optimized 3D Map Delivery System}

The present invention relates to a system for providing an optimized 3D map based on a precision location. More specifically, the present invention has a high-quality low-capacity characteristic by receiving and analyzing and rendering real image and map coordinate information, and expressing map coordinate information on a road as points and lines. The present invention relates to an accurate location based optimized 3D map providing system capable of providing a precise 3D map by aligning lanes.

First of all, unlike two-dimensional images, three-dimensional images are expressed in three dimensions using geometric data (stored using spatial coordinates with three axes of height, width, and depth). It is an image that is processed and output and expressed in three dimensions. In this case, the 3D image may be produced through a graphic process such as 3D modeling and rendering, and may be generated as a stereoscopic and realistic screen by applying optical illusions and perspectives instead of a flat and flat screen.

The 3D image may be applied to various media such as animation and software. In particular, the 3D image may be applied to a map to provide three-dimensional and realistic terrain.

At this time, the Republic of Korea Patent No. 10-0966478 (name of the invention: 3D map service providing method and geographic information system) is registered with respect to the technology for providing 3D map.

The prior art is the step of extracting a two-dimensional (2D) map image in accordance with the user's map search request in the GIS server; And providing, by the web server, a map image in 3D form using the extracted 2D map image, wherein providing the map image in 3D form comprises: extracting the 2D map image. Dividing the image into blocks of a predetermined size, performing shape and size transformation on each of the divided image blocks, and synthesizing the image blocks of which the shape and size are deformed to provide the map image in the 3D form. And performing shape and size transformation on each of the divided image blocks by reducing and deforming the top and bottom of the image block in opposite directions based on a vertical line dividing the 2D map image. Performing shape deformation to deform the shape of the image block, and for each image block of which the shape is deformed, A method of providing a 3D map service is provided, including a step of performing a size transformation to change the size of the image block whose shape is deformed according to the position of the image block based on a horizontal line dividing the 2D map image.

The prior art provides a 3D map service by transforming a 2D map image, but there is a problem that accuracy and precision of a location may be difficult in the process of rendering a 2D map to a 3D image.

Therefore, in order to solve the problems described above, after filtering and analyzing the photorealistic image, a three-dimensional image is generated, and the map coordinate information on the road is represented by points and lines to align lane positions to improve accuracy. There is a need to develop an optimized 3D map providing system.

The present invention has been made to overcome the problems of the above technology, by receiving the image and the map coordinate information received by analyzing and rendering it has a high-definition low-capacity characteristics by expressing the map coordinate information on the road in points and lines to align the lanes The main objective is to provide a precise 3D map.

Another object of the present invention is to provide a 3D map of which capacity is optimized and quality is improved by selecting a representative image in consideration of shooting conditions among a plurality of live images captured for the same area.

Still another object of the present invention is to control the transparency of image coordinates for selecting a representative image, thereby preventing the overlapping portions from being hidden.

It is a further object of the present invention to provide a higher quality 3D map by detecting, removing and correcting residual noise of a three-dimensional image.

In order to achieve the above object, according to the present invention, an accurate location-based optimized 3D map providing system includes: a data input module configured to input a photorealistic image required for map production and a plurality of map coordinate information stored for each photorealistic image from a data server; An image analysis unit for filtering and analyzing the actual image to reconstruct the intermediate image, and calculating a detail level value of the intermediate image and performing allocation processing and rendering of polygons included in the intermediate image according to the detail level value 3 An image optimization module including an optimizer configured to generate a dimensional image and store it in the data server; On a reference line which is one of a coordinate display unit for displaying the plurality of map coordinate information as a point on a road displayed in the three-dimensional image, and a plurality of lines connecting the points along an extension direction of the road. A road position correction module having a position alignment unit for aligning lane positions of roads displayed on the 3D image accordingly; And a map providing module for generating and outputting a 3D map by combining a background image based on the 3D image in which the lane positions are aligned.

The system may further include an image selection module configured to extract a specific photorealistic image according to a shooting condition from among the plurality of photorealistic images photographed for the same region, select the representative image, and store the data in the data server. It features.

In addition, the image selection module may include a plane setting unit configured to set a two-dimensional plane by defining an x-axis and a y-axis according to the shooting conditions of the live-action image with respect to the live-action image photographed in plural for the same area; A coordinate setting unit for setting image coordinates of the actual image for arranging an image on the two-dimensional plane, and a layout generating unit for generating a layout by placing the actual image on the two-dimensional plane according to the image coordinates; A matrix generator configured to determine a variance of the real image based on image coordinates, and then designate the variance as a random variable and generate a covariance matrix for each of the two random variables; the origin of the two-dimensional plane and the image A distribution that calculates a distribution weight for each real image based on the distance between coordinates and the covariance matrix. Weights and it characterized in that it comprises calculating section and, group and extracts an inspection image on the map corresponding to the reference value, weight distribution in excess of the set selected as a representative image representing the image extraction portion that stores the data server.

According to the present invention, an accurate 3D map providing system based on precision,

1) It has high quality and low capacity and can provide precise 3D map,

2) By selecting a representative image can provide a 3D map with optimized capacity and improved quality,

3) By controlling the transparency of the image coordinates for the representative image selection, not only can the overlapped portions be hidden and not visible,

4) Higher quality 3D maps can be provided by detecting, removing, and correcting residual noise in three-dimensional images.

1 is a conceptual diagram showing a schematic configuration of a system of the present invention.
2 is a block diagram showing the basic configuration of the system of the present invention.
3 is a conceptual diagram illustrating an embodiment of the filtering and analysis of the photorealistic image of the present invention.
4 is a conceptual diagram illustrating an embodiment in which a plurality of map coordinate information of the present invention is displayed with dots and lines.
5 is a conceptual diagram illustrating one embodiment of a generated 3D map of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. The accompanying drawings are not drawn to scale, and like reference numerals in each of the drawings refer to like elements.

1 is a conceptual diagram showing a schematic configuration of a system of the present invention.

Referring to FIG. 1, the precision location-based optimized 3D map providing system 1 of the present invention receives a photorealistic image and map coordinate information from a data server, generates a three-dimensional image through an analysis and rendering process, and arranges the same. And calibrate to generate an accurate 3D map.

The system 1 of the present invention performing such a function basically consists of a main server on which the system 1 of the present invention is implemented and a data server for providing and storing information and data necessary for making a map.

At this time, the main server performs a function of filtering and analyzing the live image to generate a high-quality low-capacity 3D image and storing it in a data server.

It can be understood that the main server and the data server are implemented in one system format, or the system 1 of the present invention is implemented in a cloud format on the web to analyze and optimize 3D data to generate 3D images. After that, it is also possible to store the data in a cloud data server itself, and in this case, the main server may be a server computer that manages the cloud. Preferably, the system 1 of the present invention should be implemented on the web, not in the form of a program. In the case of the main server, the system 1 should also play a role of a management server for a web page that performs processing of photorealistic images and 3D image generation.

The main server of the present invention refers to hardware having a CPU and a storage means in a state of having a communication unit and a transmission means for communicating with a data server and transmitting information, and a series of modules to be described later by software to be executed in the CPU. And specific functions thereof may be derived.

In other words, the main server is a hardware-based system having a central processing unit (CPU) and storage means such as a memory and a hard disk, that is, a software that can be executed in the central processing unit, that is, software can be installed to execute the software. A detailed configuration of the module will be described later as a structural unit called a module, a part, and an interface.

In this case, the main server may include random access memory (RAM) and read only memory (ROM), which temporarily and / or permanently store signals (or data) processed therein, It may include a processor.

In addition, the main server may be implemented in the form of a system on chip (SoC) including at least one of a graphic processor, a RAM, and a ROM.

The processor may include one or more cores (not shown) and connection passages (eg, buses) for transmitting and receiving signals to and from graphics processors (not shown) and / or other components.

The memory may store programs (one or more instructions) for execution and control of a module to a part to be described later. Programs stored in the memory may be divided into a plurality of modules according to their functions.

The steps of a method or algorithm described in connection with an embodiment of the present invention may be implemented directly in hardware, in a software module executed by hardware, or by a combination thereof. Software modules may include random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any form of computer readable recording medium well known in the art.

That is, the components of the present invention may be implemented as a program (or an application) and stored in a medium in order to be executed in combination with a computer which is hardware. The components of the present invention may be implemented as software programming or software elements, and likewise, embodiments may include various algorithms implemented in combinations of data structures, processes, routines or other programming constructs, such as C, C ++. It may be implemented in a programming or scripting language such as Java, an assembler, or the like. The functional aspects may be implemented with an algorithm running on one or more processors.

The configuration of the 'module' or 'unit' or 'interface' refers to a configuration of software such as FPGA or ASIC or software that is executed through CPU and memory while being installed and stored in the storage means of the main server. At this time, the configuration of the 'module', 'unit', or 'interface' is not limited to hardware, and may be configured to be in an addressable storage medium or may be configured to reproduce one or more processors. As an example, a 'module' or 'part' or 'interface' may refer to components such as software components, object-oriented software components, class components, and task components, processes, functions, attributes, and the like. Fields, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, data servers, data structures, tables, arrays, and variables.

The functionality provided by these 'modules' or 'parts' or 'interfaces' may be combined into a smaller number of components and 'parts' or 'modules' or by additional components and 'parts' or 'modules'. Can be further separated.

Hereinafter, detailed configurations and functions within such a macroscopic configuration will be described.

2 is a block diagram showing the basic configuration of the system of the present invention, Figure 3 is a conceptual diagram showing an embodiment of the filtering and analysis of the photorealistic image of the present invention, Figure 4 is a plurality of map coordinate information of the present invention FIG. 5 is a conceptual diagram illustrating an embodiment represented by dots and lines, and FIG. 5 is a conceptual diagram illustrating an embodiment of a generated 3D map of the present invention.

Referring to FIG. 2, the system 1 of the present invention may basically include a data input module 100, an image optimization module 200, a road position correction module 300, and a map providing module 400.

The data input module 100 provides a function of receiving a live image necessary for making a map and a plurality of map coordinate information stored for each live image from a data server.

In this case, a plurality of live-action images input from the data server may exist in any one position. For example, there may be photorealistic images taken in various external and environmental conditions, such as black and white photographs, photographs taken in backlight conditions, photographs taken on a cloudy day, and for the purpose of producing accurate and optimized 3D maps. May be provided on the system. At this time, although there may be a plurality of live image for a location, a plurality of location information may exist in one live image according to the range of the location region represented by one implementation image. In addition, the map coordinate information may be information representing a location such as latitude, longitude, altitude, detailed address, etc. of the location on the live image, and may additionally include information on the date and time at which the live image was taken. have.

The image optimization module 200 may include an image analyzer 210 and an optimizer 220.

The image analyzer 210 filters and analyzes the live image to reconstruct the intermediate image as shown in FIG. 3.

In this case, filtering refers to a process of filtering out images that are impossible to identify objects or that are not necessary for generating a 3D map, from the live image and the map coordinate information. In addition, the analysis process basically includes the identification and correction of objects and backgrounds, the separation of objects and backgrounds, and the recombination of each object, as well as the identification of objects included in the actual image. Therefore, by separating the necessary object or position information from the photo-realistic image, performing correction on the object, and performing recombination, filtering significant pixels of each pixel included in the photo-realistic image, adjusting the resolution, etc. It is also possible to perform both secondary capacity optimization and image quality optimization together. Alternatively, in addition to this, various filtering and analysis are possible, and since this can apply most of the known techniques in image and image filtering, a detailed description thereof will be omitted. The live image which has undergone such filtering and analysis may be generated as an intermediate image as shown in FIG. 3.

The optimizer 220 calculates a detail level value of the intermediate image, performs an allocation process and renders polygons included in the intermediate image according to the detail level value, and generates a 3D image and stores the same in a data server. to provide.

In this case, the polygon is a polygon of the smallest unit used to express a three-dimensional shape, that is, an object on a three-dimensional image. A polyline is formed by connecting a starting point and an ending point with a line in three-dimensional space. A straight line is then used to represent the curve and to represent the object's rim. In addition, rendering refers to a process of creating a three-dimensional image by injecting realism into the two-dimensional image in consideration of external information such as a light source, a position, and a color. As described above, since the technology for producing the 3D image by performing the allocation processing and rendering of the lolligon by the level value of the intermediate image is a known technique, a detailed description thereof will be omitted.

That is, the optimizer 220 calculates the detail level value of the intermediate image, assigns polygons according to the calculated detail level value, and then three-dimensionalizes the polygons. have. In addition, in the present invention, the 3D image is not limited to only one image or may be a concept including a plurality of image frames such as an image.

The road position correction module 300 may include a coordinate display unit 310 and a position alignment unit 320.

The coordinate display unit 310 serves to display the plurality of map coordinate information as a point on the road displayed in the 3D image, and the position alignment unit 320 may perform the point along the extension direction of the road. It provides a function to align the lane position of the road displayed on the three-dimensional image in accordance with the reference line, which is one of a plurality of lines (connected).

Here, the map coordinate information can be obtained with accurate coordinate values through Lidar (a device used to measure not only the distance to the target object but also the moving speed and direction, temperature, ambient air substance analysis and concentration measurement). When aligning the position of the three-dimensional image to be described later and create a 3D map can contribute to have a high precision characteristics.

That is, the three-dimensional image is formed by combining a plurality of real-world images after analysis, and a plurality of map coordinate information stored in each live-action image may be formed on the three-dimensional image, as shown in FIG. 4. Is indicated by a dot, and these dots are connected in particular along the extension direction of the road to form a single line, which may be formed in plural numbers depending on the width or extension direction of the road.

At this time, it is preferable to set the reference line as the baseline to be as uniform and straight as possible for ease of alignment and arrangement. For example, the line on the road that is formed in a straight line on the map coordinates may be set as a reference line, and then the positions of other lanes may be aligned based on the reference line. Here, the lane of the road may be referred to as the most monotonous and standardized structure in the photo-realistic image. As shown in FIG. 4, the lane position may be precisely arranged based on the reference line, and more accurate coordinate information may be obtained from the 3D map. It provides a foundation to reflect.

The map providing module 400 generates and outputs a 3D map by combining a background image based on the 3D image in which the lane positions are aligned. That is, the 3D map of FIG. 5 may be completed by combining background images of surrounding buildings, structures, traffic lights, etc. for the detailed completion of the 3D map to the positions aligned through the above-described position alignment unit 320. Such a background image may also be arranged and arranged in map coordinate information in the live-action image to be located at an accurate place on the 3D image.

According to such a configuration, based on map coordinate information based on a small amount of images, a feature that generates and provides a more accurate and accurate 3D map is given.

In addition, the system 1 may further include an image selection module 500.

The image selecting module 500 provides a function of extracting a specific photorealistic image from the plurality of photorealistic images photographed for the same region according to the photographing conditions, selecting the representative photographic image, and storing the data on the data server.

Here, the photorealistic images photographed in plural in the same region may be images in which the photographing conditions such as direction, brightness, illuminance, resolution, composition, perspective, exposure, and focal length are different. Each of the photorealistic images may be selected as one or a plurality of photorealistic images as representative images through a kind of process through an image selection module 500 which will be described later. When combining images, it can be a reference among multiple images, which can help optimize capacity and improve 3D map quality. A specific method of selecting the representative image will be described through a process to be described later.

At this time, the image selection module 500 is more specifically the plane setting unit 510, the coordinate setting unit 520, the arrangement map generator 530, the matrix generator 540, the distribution weight calculator 550, the representative image By including the extractor 560, it is possible to extract the representative image by systematically comparing the shooting conditions of each real image.

The plane setting unit 510 provides a function of setting a two-dimensional plane by defining an x-axis and a y-axis according to the shooting conditions of the live-action image for the live-action image photographed in plural for the same area. For example, if the shooting conditions are 'resolution' and 'luminosity', the two-dimensional plane may be set by setting the x-axis to 'resolution' and the y-axis to 'luminance'. At this time, the shooting conditions defined by the x-axis and the y-axis are preferably quantified for the placement of the live-action image, but may be arranged on the plane as an absolute unit and numerical value of the resolution and brightness of the live-action image, It is also possible to assign a numerical value relative to the shooting conditions and arrange it on the plane by reflecting it. Since this can also be identified through known image and image analysis techniques, a detailed description thereof will be omitted.

The coordinate setting unit 520 serves to set image coordinates of the actual image in order to arrange the actual image on the two-dimensional plane, and captures shooting conditions corresponding to the x and y axes defined as described above. It provides a function to numerically express the photorealistic image with respect to shooting conditions as a reference and then set it as image coordinates. For example, when the 'resolution' relative value of the live image A is 30 and the 'luminance' relative value is 10, the image coordinate of the live image A becomes (30, 10).

The layout generating unit 530 serves to generate a layout by arranging the photorealistic image on the two-dimensional plane according to the image coordinates, and the layout generated by arranging a plurality of photorealistic images corresponds to a set shooting condition. It performs a function to easily grasp the live image. In this case, the image coordinate may be given to the center of the live image to arrange the live image on the layout. In addition, the layout can also be seen that the functional superiority of the system of the present invention by helping to easily grasp how much of the photorealistic image corresponding to the shooting conditions are distributed.

와 같이 생성될 수 있다.The matrix generator 540 determines a variance of the actual image based on the image coordinates, and then provides a function of generating a covariance matrix for each of the two random variables by designating the variance as a random variable. In this case, the variance means the degree to which the real image is distributed in the layout. Since the real image is derived from the probability that the image meets the shooting conditions, the covariance matrix is distributed to the degree of dispersion of the real image (that is, how well the real image matches the shooting conditions. It can be used to determine the suitability of the photorealistic image (where fitness refers to the degree to which the photorealistic image satisfies the shooting conditions). In addition, the covariance matrix may be obtained using standard deviations of a plurality of shooting conditions, and may be used as an index for measuring how the plurality of shooting conditions affect each other. That is, the present invention may be used as an index of variation according to surrounding conditions and environments. In the present invention, two random variables are applied to a covariance matrix and compared. For example, the covariance matrix D reflecting the variance and standard deviation of the two shooting conditions of the live-action image from the layout

Can be generated as

The distribution weight calculator 550 provides a function of calculating a distribution weight for each real image based on the distance between the origin of the 2D plane and the image coordinates and the covariance matrix. In this case, the distribution weight means a value given according to the degree of distribution in the layout of the actual image. Specifically, the distribution of the live image evenly in the layout means that the live image is often generated when the system 1 of the present invention does not meet the shooting conditions, and thus the distribution weight is small. On the contrary, the uneven distribution of the live image in the layout means that the live image is often generated when the shooting conditions are satisfied. Thus, the distribution weight is large.

The representative image extractor 560 extracts a live image on a layout corresponding to a distribution weight value exceeding a preset reference value, selects a representative image, and stores the image as a representative image. For example, when the reference value is 10, it extracts a photorealistic image having a distribution weight exceeding 10 days as a representative image. In this case, since the representative image means a live image exceeding a reference value, a plurality of representative images may be extracted. Here, the reference value is a value that can be arbitrarily set by the administrator of the system 1 or automatically set by the system 1 itself, and there is no limitation in the method of setting the reference value.

In addition, the above-described distribution weight may be calculated through Equation 1 below.

Equation 1.

는 분포 가중치,

은 상기 실사 이미지

의 이미지 좌표.

는 상기 배치도의 중심 좌표,

는 상기 공분산 행렬을 의미한다.here,

Is the distribution weight,

Live action image

Image coordinates.

Is the center coordinate of the layout,

Means the covariance matrix.

Equation 1 is a formula for calculating the distribution weight for the real image based on the distance between the real image and the center of the layout and the inverse of the covariance matrix, which is not only the distance between the singular image and the center of the layout, but also the degree of dispersion of the real image. Considering the above, it is possible to accurately determine how the live image matches the shooting conditions.

In detail, if the distribution weight is calculated based only on the distance between the live image and the center of the layout, the distribution of the live image is far from the center of the layout, regardless of the values of the x and y components. The weight is largely calculated so that it is not possible to accurately determine whether or not the live image meets the shooting conditions. For example, when comparing a live image with an image coordinate of (50,50) and a live image with (-50, -50), the live image with (50.50) is more suitable for shooting conditions, but the difference between the two live image and the center of the layout Since the distances are the same, the same distribution weight is calculated. It is the covariance matrix for correcting such an error that can be accurately determined by properly considering the real image.

일 때, 실사 이미지 a에 대한 분포 가중치를 산출하면 다음과 같다.For example, the image coordinates of the live-action image a are (2,3), and the covariance matrix D is

In this case, the distribution weight for the actual image a is calculated as follows.

That is, the distribution weight of the actual image a calculated based on Equation 1 is 11.88. Through this method, the goodness of fit of the actual image can be quantified to identify the actual image that best matches the shooting condition.

In addition, when a plurality of photorealistic images are arranged on a layout plan which is one two-dimensional plane, the photorealistic images may overlap. In this case, visibility of overlapping regions (ie, portions where the photorealistic images overlap) of the photorealistic images may occur. The image selection module 500 may include a transparency control unit 570.

The transparency control unit 570 may play a role of controlling transparency of the overlapped area when the photorealistic image on the arrangement map is overlapped, and is calculated through Equation 2 below.

Equation 2.

는 투명도,

은 실사 이미지의 중첩 영역의 기존 투명도,

는 중첩된 영역의 면적,

는 실사 이미지의 픽셀 개수,

은 중첩된 영역 n의 픽셀 개수, n은 중첩된 영역에 부여된 인덱스, m은 중첩된 영역의 총 개수이며,

이다.here,

The transparency,

Is the existing transparency of the overlapping region of the photorealistic image,

Is the area of the overlapped area,

Is the number of pixels in the photorealistic image,

Is the number of pixels in the overlapped area n, n is the index given to the overlapped area, m is the total number of overlapped areas,

to be.

Equation 2 is an expression for calculating the transparency in the overlapped area based on the actual image compared to the area of the overlapped area and the number of pixels of the overlapped area, and, for example, the above-mentioned live-action image of the two overlapped photorealistic images. It adjusts the transparency of the to play the role of showing the overlap region of the photorealistic image of the lower side. In this case, when there are three or more overlapping photorealistic images, the transparency of each overlapping region may also be controlled based on Equation 2.

는 중첩 영역의 면적을 의미하는데 무차원의 값으로 중첩되는 실사 이미지의 가로 및 세로 길이를 기준으로 그 비율에 따라 산출될 수 있다.(예를 들어, 가로 길이를 50, 세로 길이를 100으로 놓고 이를 기준으로 하여 중첩 영역의 면적을 비율로 계산하여 산출할 수 있다.)In this case, transparency means the degree of transparency of the live-action image, and has a value between 0 and 1, which means that the closer to 0, the transparent, and the closer to 1, the opacity. Also,

Denotes the area of the overlapping area, which can be calculated according to the ratio based on the horizontal and vertical lengths of the photorealistic images overlapping the dimensionless values (for example, the horizontal length is 50 and the vertical length is 100. Based on this, the area of the overlapping area can be calculated by calculating the ratio.)

이 10,

이 5,

가 50이라고 하면 중첩 영역의 투명도

는 수학식 2에 의해 다음과 같이 산출된다.As a numerical example for the above Equation 2, when two photorealistic images overlap, the existing transparency of the overlapped region is 0.8,

This 10,

This 5,

Is 50, the transparency of the overlapping region

Is calculated by Equation 2 as follows.

That is, the transparency of the overlapped region calculated through Equation 2 is 0.24, so that the overlapped portion may be covered by the transparent region.

In this way, a layout may be generated based on the photorealistic image, and the portion where the photorealistic image overlaps may be controlled by managing transparency to provide a more realistic photorealistic image.

In another embodiment, the image optimization module 200 may further include an image corrector 230 to correct a noise that may remain in the live image to provide a higher quality 3D map.

The image corrector 230 detects and removes residual noise from the generated 3D image to perform correction on the 3D image.

That is, a higher quality 3D image may be provided by optimizing the real image and then rendering and correcting the residual noise that may occur in the process of generating the 3D image.
As described so far, the precise location-based optimized 3D map providing system according to the present invention has been represented in the above description and drawings, but this is merely an example, and the spirit of the present invention is not limited to the above description and drawings. Various changes and modifications are possible without departing from the spirit of the invention.

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

delete

1: system 100: data input module
200: image optimization module 210: image analysis unit
220: optimizer 230: image correction unit
300: road position correction module
310: coordinate display unit 320: position alignment unit
400: map providing module 500: image selection module
510: plane setting unit 520: coordinate setting unit
530: layout generator 540: matrix generator
550: Distribution weight calculation unit 560: Representative image extraction unit
570: transparency control

Claims (12)

Precise location based optimized 3D map providing system,
A data input module for receiving a live image necessary for making a map and a plurality of map coordinate information stored for each live image from a data server;
An image analysis unit for filtering and analyzing the actual image to reconstruct the intermediate image, and calculating a detail level value of the intermediate image and performing allocation processing and rendering of polygons included in the intermediate image according to the detail level value 3 An image optimization module including an optimizer configured to generate a dimensional image and store it in the data server;
On a reference line which is one of a coordinate display unit for displaying the plurality of map coordinate information as a point on a road displayed in the three-dimensional image, and a plurality of lines connecting the points along an extension direction of the road. A road position correction module having a position alignment unit for aligning lane positions of roads displayed on the 3D image accordingly;
And a map providing module for generating and outputting a 3D map by combining a background image on the basis of the 3D image in which the lane positions are aligned. 3. The method of claim 1,
The system,
And an image selection module for extracting a specific photorealistic image from the plurality of photorealistic images photographed for the same region according to shooting conditions, selecting the representative photographic image, and storing the image in the data server. Map providing system. The method of claim 2,
The image selection module,
A plane setting unit configured to set a two-dimensional plane by defining an x-axis and a y-axis according to the shooting conditions of the live-action image with respect to the live-action image photographed in plural for the same area;
A coordinate setting unit for setting image coordinates of the actual image to arrange the actual image on the two-dimensional plane;
A layout generating unit for generating a layout by arranging the photo-realistic image on the two-dimensional plane according to the image coordinates;
A matrix generator configured to determine a variance of the actual image based on the image coordinates, and then designate the variance as a random variable and generate a covariance matrix for each of the random variables;
A distribution weight calculator configured to calculate a distribution weight for each real image based on a distance between the origin of the two-dimensional plane and the image coordinates and the covariance matrix;
And a representative image extracting unit for extracting a photorealistic image corresponding to a distribution weight value exceeding a preset reference value, selecting the representative image as a representative image, and storing the extracted representative image in the data server. The method of claim 3, wherein
The distribution weight is,
Optimized 3D map providing system, characterized in that calculated through the following equation (1).
Equation 1.

(here,

Is the distribution weight,

Live action image

Image coordinates.

Is the center coordinate of the layout,

Is the covariance matrix) The method of claim 3, wherein
The image selection module,
And a transparency controller configured to control transparency of the overlapped area when two photorealistic images overlap each other on the layout. The method of claim 5,
The transparency is,
Optimized 3D map providing system, characterized in that it is calculated through the following equation (2).
Equation 2.

(here,

The transparency,

Is the existing transparency of an area where two photorealistic images overlap each other,

Is the area of the overlapped area,

Is the number of pixels in the photorealistic image,

Is the number of pixels of the overlapped area n, n is the index given to the overlapped area, m is the total number of the overlapped areas,

to be.) The method of claim 1,
The image optimization module,
And an image correction unit configured to detect and remove residual noise from the generated 3D image to perform correction on the 3D image. delete delete delete delete delete

Images