JP2001091251A - Three dimensional map preparing method - Google Patents
Three dimensional map preparing methodInfo
- Publication number
- JP2001091251A JP2001091251A JP26812199A JP26812199A JP2001091251A JP 2001091251 A JP2001091251 A JP 2001091251A JP 26812199 A JP26812199 A JP 26812199A JP 26812199 A JP26812199 A JP 26812199A JP 2001091251 A JP2001091251 A JP 2001091251A
- Authority
- JP
- Japan
- Prior art keywords
- building
- dimensional map
- height
- roof
- correlation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Measurement Of Optical Distance (AREA)
- Processing Or Creating Images (AREA)
Abstract
Description
【０００１】[0001]
【発明の属する技術分野】この発明は、複数の航空写真
を利用した３次元地図作成方法及び装置の技術分野に属
する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention belongs to the technical field of a method and an apparatus for creating a three-dimensional map using a plurality of aerial photographs.
【０００２】[0002]
【従来の技術】３次元地図は地上の建造物を立体的に表
示し、街の外観等をより現実的に表現し、分かり易くす
ることが３次元地図作製の目的の１つである。従って、
３次元地図をより正確に作成するためには建造物の地上
からの高さは、外観のテクスチャと共に重要な情報であ
る。また、特定の形状の屋根を持つ建造物は目印となる
から、これを３次元地図に表すことは意義のあることで
ある。2. Description of the Related Art One of the objectives of creating a three-dimensional map is to three-dimensionally display a building on the ground in a three-dimensional manner, to render the appearance of the city more realistic, and to make it easier to understand. Therefore,
In order to create a three-dimensional map more accurately, the height of the building from the ground is important information together with the appearance texture. In addition, since a building having a roof having a specific shape is a landmark, it is meaningful to represent this on a three-dimensional map.
【０００３】しかし、航空写真から３次元地図を作成す
る技術は近年に至って研究が進み始めた状況で、安価で
実用的に利用できる技術は公表されていない。本出願人
は安価で実用的な作成方法について出願を行った（平成
１１年特許願第１４０１４３号）。本出願の発明は前記
出願に記載された技術を改良したものである。即ち、前
記出願においては建物の高さを求めるための計算時間が
長く課題であった。また、屋根形状を平屋根に限定して
いるため、特徴を持った建物の表示方法に制限を受け、
このためテクスチャの張り付けの場合の領域の決定方法
等にも課題があった。このような課題を解決するために
本発明がなされた。[0003] However, the technology for creating a three-dimensional map from an aerial photograph has recently been researched, and no technology that can be used inexpensively and practically has been published. The present applicant has applied for an inexpensive and practical production method (Japanese Patent Application No. 140143/1999). The invention of the present application is an improvement of the technology described in the aforementioned application. That is, in the above-mentioned application, the calculation time for obtaining the height of the building was long and was a problem. In addition, since the roof shape is limited to a flat roof, the display method of buildings with features is limited,
For this reason, there is also a problem in a method of determining an area in the case of attaching a texture. The present invention has been made to solve such a problem.
【０００４】[0004]
【発明が解決しようとする課題】この発明は、上述のよ
うな背景の下になされたもので、建物の高さを短時間で
計算でき、しかも平屋根以外の屋根形状を持つ建物にも
適用できる３次元地図作成方法を提供することを課題と
している。SUMMARY OF THE INVENTION The present invention has been made under the above background, and can be applied to a building having a roof shape other than a flat roof, in which the height of the building can be calculated in a short time. It is an object of the present invention to provide a three-dimensional map creation method that can be used.
【０００５】[0005]
【課題を解決するための手段】上記課題を解決するため
に本発明は以下の手段を採用している。即ち、請求項１
記載の発明は、複数の航空写真を利用した３次元地図作
成方法において、地上の建物位置を示す２次元地図に射
影変換行列を作用させて、該航空写真における建物の地
上の位置座標を求める工程を含むことを特徴としてい
る。The present invention employs the following means to solve the above-mentioned problems. That is, claim 1
According to the described invention, in a three-dimensional map creating method using a plurality of aerial photographs, a step of applying a projective transformation matrix to a two-dimensional map indicating the positions of buildings on the ground to obtain position coordinates of the buildings in the aerial photographs on the ground. It is characterized by including.
【０００６】請求項２記載の発明は、複数の航空写真を
利用した３次元地図作成方法において、該３次元地図作
成方法は該航空写真を利用して建物の高さを求める工程
を含み、該工程は建物の高さを複数レベル仮定し、仮定
された任意の高さについて該航空写真間の該建物のテキ
スチャの相関を求め、相関が最大となる建物の高さを該
建物の高さと決定するようにしたことを特徴としてい
る。According to a second aspect of the present invention, there is provided a three-dimensional map creating method using a plurality of aerial photographs, wherein the three-dimensional map creating method includes a step of obtaining a building height using the aerial photographs. The process assumes a plurality of levels of building heights, finds a correlation of the texture of the building between the aerial photographs for an assumed height, and determines the height of the building having the largest correlation as the height of the building. It is characterized by doing so.
【０００７】請求項３記載の発明は、複数の航空写真を
利用した３次元地図作成方法において、該３次元地図作
成方法は該航空写真を利用して建物の高さを求める工程
を含み、該工程は建物の高さを複数レベル仮定し、仮定
された任意の高さについて該航空写真間の該建物のテキ
スチャの相関を求め、相関が急激に変化する建物の高さ
を該建物の高さと決定するようにしたことを特徴として
いる。According to a third aspect of the present invention, there is provided a method for creating a three-dimensional map using a plurality of aerial photographs, the method comprising the step of determining the height of a building using the aerial photographs. The process assumes a plurality of levels of building heights, finds a correlation of the texture of the building between the aerial photographs for any assumed height, and determines the height of the building where the correlation changes rapidly with the height of the building. It is characterized in that it is determined.
【０００８】請求項４記載の発明は、請求項１〜請求項
３に記載の発明において、前記建物の高さを求める工程
は、建物のテキスチャの相関を求める前に、建物の形状
を一致させる変換を行う工程を含むことを特徴としてい
る。According to a fourth aspect of the present invention, in the first aspect of the present invention, the step of obtaining the height of the building matches the shape of the building before obtaining the correlation of the texture of the building. It is characterized by including a step of performing conversion.
【０００９】請求項５記載の発明は、請求項１〜請求項
４に記載の発明において、前記建物の高さを求める工程
は、建物のテキスチャの相関を求める前に、建物の形状
を一致させる変換を行う工程を含むことを特徴としてい
る。According to a fifth aspect of the present invention, in the first aspect of the present invention, the step of obtaining the height of the building matches the shape of the building before obtaining the correlation of the texture of the building. It is characterized by including a step of performing conversion.
【００１０】請求項６記載の発明は、請求項１〜請求項
５に記載の発明において、前記建物の高さを求める工程
は、該建物の屋根の形状を複数のパターンから選択し、
選択された屋根パターンについて相関を求める両者の航
空写真で共通に見える領域を求め、該領域についてテキ
スチャの相関を求めるようにしたことを特徴としてい
る。また、請求項７記載の発明は、請求項６に記載の発
明において、前記屋根パターンはドーム形屋根、４角錐
形屋根、平屋根を含むことを特徴としている。According to a sixth aspect of the present invention, in the first aspect of the present invention, the step of obtaining the height of the building includes selecting a shape of a roof of the building from a plurality of patterns,
It is characterized in that an area that looks common in both aerial photographs for obtaining a correlation for the selected roof pattern is obtained, and a texture correlation is obtained for the area. The invention according to claim 7 is the invention according to claim 6, wherein the roof pattern includes a dome-shaped roof, a pyramid-shaped roof, and a flat roof.
【００１１】請求項８記載の発明は、請求項１〜請求項
７に記載の発明において、前記３次元地図作成方法は、
地上面が傾斜又は平坦でない場合は、任意高さの水平面
に２次元地図を作製し、該水平面を基準として建物の高
さを求めるようにしたことを特徴としている。[0011] The invention according to claim 8 is the invention according to claims 1 to 7, wherein the three-dimensional map creation method comprises:
When the ground surface is not inclined or flat, a two-dimensional map is created on a horizontal plane of an arbitrary height, and the height of the building is determined based on the horizontal plane.
【００１２】請求項９記載の発明は、請求項１〜請求項
８に記載の発明において、前記３次元地図作成方法は、
２次元地図に含まれる建物の平面形状がポリゴンでない
建物は複数のポリゴンに分割して３次元地図を作製する
ことを特徴としている。According to a ninth aspect of the present invention, in the first aspect of the present invention, the three-dimensional map creating method includes:
A feature is that a building included in a two-dimensional map and whose planar shape is not a polygon is divided into a plurality of polygons to create a three-dimensional map.
【００１３】[0013]
【発明の実施形態】図２は本発明の実施形態における全
体の手順の流れを示す。図３は２次元地図を利用して航
空写真における建物の地面位置の座標を求める方法を示
す。図４は相関の計算を容易にするために建物の方向を
変換した例を示す。図１は本発明における平屋根の場合
の建物高さを求める基本原理を示す。以下、図面を参照
してこの発明の実施形態について説明する。FIG. 2 shows an overall procedure flow in an embodiment of the present invention. FIG. 3 shows a method of obtaining the coordinates of the ground position of a building in an aerial photograph using a two-dimensional map. FIG. 4 shows an example in which the direction of the building is changed to facilitate the calculation of the correlation. FIG. 1 shows a basic principle for obtaining a building height in the case of a flat roof according to the present invention. Hereinafter, embodiments of the present invention will be described with reference to the drawings.
【００１４】図２において、ステップＳ１では複数の航
空写真を撮影する。３次元地図を作製する道路や建物等
を含んだ航空写真を異なる位置及び角度から撮影する。
撮影は少なくとも２枚以上の航空写真が必要であり、色
々な角度から撮影すればより正確な３次元地図の作製が
可能になる。説明を簡単にするために、以後航空写真は
２枚、航空写真１と航空写真２の場合について説明す
る。撮影カメラは同一のカメラを使用するのが便利であ
る。なお、航空写真には３次元座標（Ｘ，Ｙ，Ｚ）の既
知である基準点が少なくとも６個以上含ませる必要があ
る。基準点としては建物の角点等の特徴のある点又は特
別に設けた目印点等を使用してもよい。In FIG. 2, in step S1, a plurality of aerial photographs are taken. Aerial photographs including roads, buildings, and the like for creating a three-dimensional map are taken from different positions and angles.
At least two or more aerial photographs are required for photographing, and a more accurate three-dimensional map can be created by photographing from various angles. For the sake of simplicity, the case of two aerial photographs, aerial photograph 1 and aerial photograph 2, will be described below. It is convenient to use the same shooting camera. Note that the aerial photograph must include at least six or more reference points whose three-dimensional coordinates (X, Y, Z) are known. As the reference point, a characteristic point such as a corner point of a building or a specially provided landmark point may be used.
【００１５】ステップＳ２では、３次元空間の点から各
航空写真の点への変換を表す射影変換行列Ｐを求める。
射影変換行列Ｐは個々の航空写真毎に異なり、射影変換
行列Ｐは（４×３）行列で１２個の未知数を含む。この
未知数を求めるためには同一平面上にない６個の基準と
なる点を用いる。キャリブレーションの精度を上げるに
は６個以上の多数の基準点を用い、最小二乗法によって
パラメータを同定する。６個以上の基準点の座標を利用
して射影変換行列を求める方法は公知であり、公知の従
来技術を利用する。In step S2, a projection transformation matrix P representing transformation from points in the three-dimensional space to points in each aerial photograph is obtained.
The projective transformation matrix P is different for each aerial photograph, and the projective transformation matrix P is a (4 × 3) matrix including 12 unknowns. In order to obtain the unknown, six reference points that are not on the same plane are used. In order to improve the accuracy of the calibration, a large number of reference points of six or more are used, and parameters are identified by the least square method. A method of obtaining a projective transformation matrix using the coordinates of six or more reference points is known, and a known conventional technique is used.
【００１６】ステップＳ３では、建物等の位置の座標が
既知である２次元地図を利用して、ステップＳ２で求め
た射影変換行列Ｐにより建物等の地面位置座標を航空写
真上で求める。２次元地図は建物、道路等の地面におけ
る位置座標関係を数値データにより明示した地図を利用
する。２次元地図は高さ方向の長さをゼロにした（Ｚ＝
０）３次元地図であり、これに射影変換行列Ｐを作用さ
せて対応点を求めれば、航空写真における建物の地面位
置座標が求められる。図３は２次元地図を航空写真１、
航空写真２に変換した例を示している。以下、航空写真
に写像された建物の２次元地図（ｐ１ｐ２ｐ３ｐ４）を
建物の地面地図と呼ぶ。なお、図３で、２次元地図の点
Ｐ１〜点Ｐ４の座標（Ｘ、Ｙ、０）は既知であるとし、
このデータを積極的に利用する。他の建物等についても
同様に地面地図を求める。In step S3, the coordinates of the ground position of the building or the like are obtained on the aerial photograph using the two-dimensional map in which the coordinates of the position of the building or the like are known using the projection transformation matrix P obtained in step S2. The two-dimensional map uses a map in which positional coordinates on the ground, such as buildings and roads, are specified by numerical data. In the two-dimensional map, the length in the height direction is set to zero (Z =
0) The map is a three-dimensional map, and when a corresponding point is obtained by applying a projective transformation matrix P to the map, the ground position coordinates of the building in the aerial photograph can be obtained. Figure 3 shows a two-dimensional map of aerial photograph 1,
The example which converted into the aerial photograph 2 is shown. Hereinafter, the two-dimensional map (p1p2p3p4) of the building mapped to the aerial photograph is referred to as a ground map of the building. In FIG. 3, it is assumed that the coordinates (X, Y, 0) of the points P1 to P4 on the two-dimensional map are known,
Actively use this data. A ground map is similarly obtained for other buildings and the like.
【００１７】ステップＳ４では建物の屋根形状を指定す
る。ここでは、屋根は平らな平屋根形状とし、それ以外
の形状の屋根については後で詳細に説明する。また、建
物等の３次元空間における形状は、２次元地図に示され
た形状（例えば図２の２次元地図における矩形（Ｐ１Ｐ
２Ｐ３Ｐ４））を上方向に垂直に伸ばし、そこに屋根が
付いた形状を仮定する。建物は通常地面から垂直に建っ
ているのでこの仮定は許容される。斜塔のような建物は
別個に処理すればよい。In step S4, the roof shape of the building is designated. Here, the roof has a flat flat roof shape, and other roof shapes will be described later in detail. Further, the shape of the building or the like in the three-dimensional space is the shape shown in the two-dimensional map (for example, a rectangle (P1P
2P3P4)) extends vertically upward, and assumes a shape with a roof. This assumption is acceptable because buildings are usually built vertically from the ground. Buildings such as leaning towers may be treated separately.
【００１８】ステップＳ５では相関法を利用して建物の
高さを求める。まず、相関を計算する前に建物の向きを
一致させる。図４の上欄に示すように、水平面地図は航
空写真毎に方向及び向きが異なるので、図４の下欄に示
すようにホモグラフィー変換により方向及び向きを一致
させる。図４では航空写真２のみを方向変換している
が、両者の方向を変換して建物の地面地図の辺をｘ軸又
はｙ軸に平行になるようにしてもよい。In step S5, the height of the building is obtained by using the correlation method. First, the directions of the buildings are matched before calculating the correlation. As shown in the upper column of FIG. 4, the horizontal plane map has different directions and directions for each aerial photograph, and thus the directions and directions are matched by homography conversion as shown in the lower column of FIG. In FIG. 4, only the direction of the aerial photograph 2 is changed, but the directions may be changed so that the side of the ground map of the building is parallel to the x-axis or the y-axis.
【００１９】屋根形状は水平面で、かつ地面地図と同形
であると仮定している。航空写真１と航空写真２の屋根
のテキスチャについて相関は以下のようにして求める。
なお、相関から建物高さを決定するためには選択された
航空写真が何れも屋根のテキスチャを完全に表している
ことが必要である。一方が他の建物等の陰に隠れて屋根
の部分が見えない場合は他の航空写真を選択する。図１
に示すように、各航空写真の建物の高さの下限及び上限
を定めて、この間を複数の高さレベルに区切る。ここで
は下限をゼロ高さとし、上限をｈmaxとしている。各高
さレベル（例えば、図のｈi)について航空写真１と２の
矩形（ＡＢＣＤ）と矩形（Ａ’Ｂ’Ｃ’Ｄ’）のテキス
チャの相関を求める。相関は両写真の対応する点のテキ
スチャのピクセルの輝度差の自乗（又は絶対値）を求め
て積分する。なお、相関を求める方法には、例えば、誤
差の二乗和を求めるＳＳＤ法、ゼロ平均正規化誤差の二
乗和を求めるＺＳＳＤ法又はゼロ平均化誤差の二乗和を
求めるＺＮＳＳＤ法等が知られている。It is assumed that the roof shape is a horizontal plane and has the same shape as the ground map. The correlation between the roof textures of the aerial photograph 1 and the aerial photograph 2 is obtained as follows.
Note that in order to determine the building height from the correlation, it is necessary that all the selected aerial photographs completely represent the texture of the roof. If one is hidden behind another building and the roof is not visible, another aerial photograph is selected. FIG.
As shown in (1), a lower limit and an upper limit of the height of the building of each aerial photograph are determined, and this is divided into a plurality of height levels. Here, the lower limit is set to zero height, and the upper limit is set to hmax. For each height level (eg, hi in the figure), the correlation between the rectangles (ABCD) of the aerial photographs 1 and 2 and the texture of the rectangles (A'B'C'D ') is determined. The correlation is integrated by finding the square (or absolute value) of the luminance difference between the pixels of the texture at the corresponding points in both photographs. In addition, as a method for obtaining the correlation, for example, an SSD method for obtaining the sum of squares of the error, a ZSSD method for obtaining the sum of squares of the zero-mean normalization error, and a ZNSSD method for obtaining the sum of squares of the zero-average error are known. .
【００２０】相関が最大になる高さレベルをその建物の
高さと決定する。これは航空写真においては、平屋根部
分は撮影角度によってテキスチャがあまり変化しない
が、外壁部分は撮影角度によってテキスチャが変化して
撮影されるために同じ部分の輝度が一致せず、相関が小
さくなる。特に、一方の航空写真に外壁の一部が見えな
い場合は相関は著しく小さくなる。また、高さレベルが
建物高さを超えた場合は背景は撮影する角度によって異
なり、相関は急激に小さくなる。従って、相関が急激に
小さくなる高さレベルを建物の高さと決定してもよい。
なお、建物の高さを求める際にポリゴン形状をしていな
い建物は仮想的に分割してポリゴン化しておいてもよ
い。ポリゴン化すると相関の計算やテキスチャのワーピ
ング等の処理が容易になる。The height level at which the correlation is maximized is determined as the height of the building. This is because, in aerial photography, the texture of the flat roof part does not change much depending on the shooting angle, but the texture of the outer wall part changes with the shooting angle, so the brightness of the same part does not match and the correlation decreases . In particular, when a part of the outer wall is not visible in one aerial photograph, the correlation becomes extremely small. When the height level exceeds the height of the building, the background differs depending on the shooting angle, and the correlation sharply decreases. Therefore, the height level at which the correlation sharply decreases may be determined as the building height.
When calculating the height of a building, a building that does not have a polygon shape may be virtually divided into polygons. Polygoning facilitates processing such as correlation calculation and texture warping.
【００２１】ステップＳ６では建物の３次元モデルを作
成する。即ち、ステップＳ３で建物の側面の形状が決定
され、ステップＳ４で屋根形状が決定され、ステップＳ
５で高さが決定されているので、これらのデータに基づ
いて建物の３次元モデルを作成する。ステップＳ７では
３次元モデルに張り付けるテキスチャを選択する。選択
は撮影した航空写真の中でその部分を最も鮮明に表して
いるテキスチャを選択するのが好ましい。従って、２枚
以上の航空写真に張り付ける部分の写真がある場合は表
示している面積最大の航空写真を選択してもよい。な
お、鮮明なテキスチャを張付けるために地上から撮影し
た写真のテキスチャを張り付けてもよい。In step S6, a three-dimensional model of the building is created. That is, the shape of the side of the building is determined in step S3, and the roof shape is determined in step S4.
Since the height is determined in 5, a three-dimensional model of the building is created based on these data. In step S7, a texture to be attached to the three-dimensional model is selected. For the selection, it is preferable to select a texture that most clearly represents the portion in the photographed aerial photograph. Therefore, when there is a photograph of a portion to be attached to two or more aerial photographs, the aerial photograph having the largest displayed area may be selected. Note that a texture of a photograph taken from the ground may be attached in order to attach a clear texture.
【００２２】ステップＳ８は選択されたテキスチャを３
次元モデルに逆ワーピングする。逆ワーピングすること
により一様なテキスチャの張り付けが行われる。逆ワー
ピングの方法は、本出願人の先出願（平成１１年特許願
第１４０１４３号）にも説明されており、本発明の要旨
ではないので詳細な説明は省略する。In step S8, the selected texture is set to 3
Reverse warping to a dimensional model. By performing the reverse warping, the texture is stuck uniformly. The reverse warping method is also described in the earlier application filed by the present applicant (Japanese Patent Application No. 140143/1999) and is not the gist of the present invention, so that the detailed description is omitted.
【００２３】なお、上記の３次元地図作製の方法では建
物等の建っている地上面は水平であると仮定している。
地上面が水平でない場合、例えば、傾斜している場合又
は凹凸がある場合は以下の予備処理を行う。図５（Ａ）
に示すように傾斜や凹凸が緩やかな場合は、１個の水平
基準面２１を定め、２次元地図を水平基準面２１で表現
し、上記した３次元地図作成方法で水平基準面２１より
上の地図を作製し、水平基準面２１より下の地図は建物
等は下側に延長し、別途に作成し、両者を合成すればよ
い。また、図５（Ｂ）に示すように傾斜や凹凸が急な場
合は、複数の水平基準面からなる基準面２２を用いて同
様な手続きを行う。In the above three-dimensional map making method, it is assumed that the ground surface on which a building or the like is built is horizontal.
When the ground surface is not horizontal, for example, when it is inclined or has irregularities, the following preliminary processing is performed. FIG. 5 (A)
In the case where the inclination or unevenness is gentle as shown in (1), one horizontal reference plane 21 is defined, the two-dimensional map is represented by the horizontal reference plane 21, and the two-dimensional map is expressed by the above-described three-dimensional map creation method. A map may be created, and a map below the horizontal reference plane 21 may be created separately by extending a building or the like downward, and then combining the two. In the case where the inclination or unevenness is steep as shown in FIG. 5B, a similar procedure is performed using the reference plane 22 including a plurality of horizontal reference planes.
【００２４】図６は建物の屋根形状のパターンの例を示
す図である。ステップＳ４で建物の屋根形状を指定する
には航空写真参照しながらパターンの中から最も近い形
状を選択し、２次元地図のその建物のデータと一緒に記
憶する。選択された形状と２次元地図のデータから建物
の屋根形状及び寸法が定まる。なお、屋根形状は寸法比
が完全に正しくなくとも近似していれば建物の外観とし
ては十分に満足するものが得られる。従って、屋根形状
のパターンは適宜に定め、必要に応じて追加可能にして
おくと便利である。選択された屋根形状パターンは２次
元地図のデータと共に建物の高さを決定するために利用
される。なお、平屋根をデフォルト値として定め、屋根
形状パターンが選択されなかった建物は平屋根とするよ
うに定めておくと便利である。FIG. 6 is a diagram showing an example of a roof-shaped pattern of a building. To specify the roof shape of the building in step S4, the closest shape is selected from the patterns while referring to the aerial photograph, and stored together with the data of the building on the two-dimensional map. The roof shape and dimensions of the building are determined from the selected shape and the data of the two-dimensional map. It should be noted that if the roof shape is similar even if the dimensional ratio is not completely correct, a satisfactory appearance can be obtained for the building. Therefore, it is convenient that the roof-shaped pattern is appropriately determined and can be added as needed. The selected roof shape pattern is used together with the data of the two-dimensional map to determine the height of the building. It is convenient to determine a flat roof as a default value and to define a flat roof for a building whose roof shape pattern is not selected.
【００２５】図７は屋根形状が４角錐の場合について建
物高さを定める手順について説明した図である。図７
（Ａ）のような屋根形状の場合には前述した平屋根の場
合の建物高さを決定する方法では正しく決定できない場
合がある。航空写真の撮影角度が異なるために一方の航
空写真では見えるが、他方の航空写真には見えない屋根
の部分が存在する。例えば、図（Ｂ）では３角形（ＰＢ
Ｃ）は見えるが、図（Ｃ）では３角形（ＰＢＣ）は裏に
隠れて見えない。従って、このまま相関を計算しても正
しい建物の高さを決定することはできない。なお、平屋
根の場合は他の建物等に遮られない限りこの様なことは
起こらない。FIG. 7 is a diagram for explaining the procedure for determining the building height when the roof shape is a quadrangular pyramid. FIG.
In the case of a roof shape as in (A), the above-described method of determining the building height in the case of a flat roof may not be correctly determined. Due to the different angles of the aerial photographs, there is a part of the roof that is visible in one aerial photograph but not visible in the other aerial photograph. For example, in FIG.
C) is visible, but the triangle (PBC) is hidden behind and cannot be seen in FIG. Therefore, even if the correlation is calculated as it is, the correct building height cannot be determined. In the case of a flat roof, this does not occur unless it is blocked by another building.
【００２６】図７（Ｄ）に示すように、航空写真で屋根
が共通に見える部分を決定する。屋根の各部分が共通し
て見えているかどうかを調べるためにはパターンマッチ
ングの技術を使用してもよいし、又は他の方法、例え
ば、その面からランダムに光線を発してその光線が写真
画面に到達するか否かを調べる光線追跡法を利用しても
よい。この様にして得られた共通に見える部分、例え
ば、図７（Ｄ）における斜線を施した３角形（ＰＡＢ）
に対して、ステップＳ５で説明した方法を適用する。即
ち、航空写真１と航空写真２の３角形（ＰＡＢ）に相当
する領域について高さレベルを変えながら相関を計算す
ることにより、建物高さを求めることができる。As shown in FIG. 7D, a portion where the roof looks common in the aerial photograph is determined. Pattern matching techniques may be used to determine if each part of the roof is visible in common, or other methods, such as randomly emitting light from the surface and the light A ray tracing method for checking whether or not the light beam arrives may be used. The commonly visible portion obtained in this manner, for example, a hatched triangle (PAB) in FIG.
, The method described in step S5 is applied. That is, the building height can be obtained by calculating the correlation while changing the height level for the area corresponding to the triangle (PAB) of the aerial photograph 1 and the aerial photograph 2.
【００２７】以上に述べた方法によれば、建物の高さや
屋根形状についても正しい３次元モデルが得られる。ま
た、建物の高さを求める相関の計算は建物の屋根に相当
する部分についてのみ相関を計算しており、相関を計算
時間が短縮できる。また、屋根形状について十分なパタ
ーンを用意しておけば、実際に近似した建物の３次元モ
デルが構成できるためより優れた３次元地図が作製でき
る。According to the method described above, a correct three-dimensional model can be obtained for the building height and roof shape. Further, the calculation of the correlation for obtaining the height of the building is performed only for the portion corresponding to the roof of the building, and the calculation time of the correlation can be reduced. In addition, if a sufficient pattern is prepared for the roof shape, a more excellent three-dimensional map can be produced because a three-dimensional model of a building that is actually approximated can be constructed.
【００２８】以上、この発明の実施形態、実施例を図面
により詳述してきたが、具体的な構成はこの実施例に限
られるものではなく、この発明の要旨を逸脱しない範囲
の設計の変更等があってもこの発明に含まれる。例え
ば、射影変換行列Ｐのキャリブレーションは実施形態で
述べた方法に限定されず、他の方法を使用してもよい。
また、航空写真は２枚に限らず、更に多数枚の航空写真
を利用してもよい。The embodiments and examples of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to the examples, and changes in the design and the like may be made without departing from the gist of the present invention. Even if there is, it is included in the present invention. For example, the calibration of the projection transformation matrix P is not limited to the method described in the embodiment, and another method may be used.
The number of aerial photographs is not limited to two, and a larger number of aerial photographs may be used.
【００２９】[0029]
【発明の効果】以上説明したように、この発明の構成に
よれば、２次元地図を利用し、建物の形状についても実
用的に許容できる範囲の妥当な仮定を設けて、相関を利
用しているので建物等の高さを求める計算時間を短縮で
きるという効果が得られる。また、屋根形状について十
分なパターンを用意しておけば、実際に近似した建物の
３次元モデルが構成できるためより優れた３次元地図が
作製できるという効果が得られる。As described above, according to the configuration of the present invention, a two-dimensional map is used, a reasonable assumption of a practically acceptable range is provided for the shape of a building, and a correlation is used by utilizing the correlation. Therefore, the effect that the calculation time for obtaining the height of the building or the like can be reduced can be obtained. Also, if a sufficient pattern is prepared for the roof shape, a three-dimensional model of a building that is actually approximated can be constructed, so that there is an effect that a more excellent three-dimensional map can be produced.
【図１】 本発明における平屋根の場合の建物高さを求
める基本原理を示す。FIG. 1 shows a basic principle for obtaining a building height in the case of a flat roof according to the present invention.
【図２】 本発明の実施形態における全体の手順の流れ
を示す。FIG. 2 shows an overall procedure flow in the embodiment of the present invention.
【図３】 ２次元地図を利用して航空写真における建物
の地面位置の座標を求める方法を示す。FIG. 3 shows a method of obtaining the coordinates of the ground position of a building in an aerial photograph using a two-dimensional map.
【図４】 相関の計算を容易にするために建物の方向を
変換した例を示す。FIG. 4 shows an example in which the direction of a building is changed to facilitate calculation of a correlation.
【図５】 地上面が水平又は平坦でない場合を説明した
図である。FIG. 5 is a diagram illustrating a case where the ground surface is not horizontal or flat.
【図６】 建物の屋根形状のパターンの例を示す図であ
る。FIG. 6 is a diagram showing an example of a roof-shaped pattern of a building.
【図７】 屋根形状が４角錐の場合について建物高さを
定める手順について説明した図である。FIG. 7 is a diagram illustrating a procedure for determining a building height when the roof shape is a quadrangular pyramid.
ＡＢＣＤ 屋根の地図 ｐ１ｐ２ｐ３ｐ４ 地面地図 ABCD Roof Map p1p2p3p4 Ground Map
フロントページの続き (72)発明者 市橋 敬男 東京都渋谷区広尾５−19−９広尾ＯＮビル 株式会社ゲン・テック内 (72)発明者 鈴木 一史 東京都渋谷区広尾５−19−９広尾ＯＮビル 株式会社ゲン・テック内 Ｆターム(参考） 2F112 AC02 AC06 BA05 CA08 CA14 FA03 FA21 FA38 FA39 FA41 5B050 BA04 BA09 BA17 EA28 FA06Continuation of the front page (72) Inventor Takao Ichihashi 5-19-9 Hiroo ON Building, Hiroo 5-19-9, Shibuya-ku, Tokyo (72) Inventor Kazushi Suzuki 5-19-9 Hiroo ON Hiroo, Shibuya-ku, Tokyo Building Gen-Tech Co., Ltd. F-term (reference) 2F112 AC02 AC06 BA05 CA08 CA14 FA03 FA21 FA38 FA39 FA41 5B050 BA04 BA09 BA17 EA28 FA06
Claims (9)
成方法において、地上の建物位置を示す２次元地図に射
影変換行列を作用させて、該航空写真における建物の地
上の位置座標を求める工程を含むことを特徴とする３次
元地図作成方法。1. A method for creating a three-dimensional map using a plurality of aerial photographs, wherein a projective transformation matrix is applied to a two-dimensional map indicating building positions on the ground to obtain position coordinates of the buildings on the aerial photograph on the ground. A three-dimensional map creation method, comprising:
成方法において、該３次元地図作成方法は該航空写真を
利用して建物の高さを求める工程を含み、該工程は建物
の高さを複数レベル仮定し、仮定された任意の高さにつ
いて該航空写真間の該建物のテキスチャの相関を求め、
相関が最大となる建物の高さを該建物の高さと決定する
ようにしたことを特徴とする３次元地図作成方法。2. A method for creating a three-dimensional map using a plurality of aerial photographs, the method comprising the step of determining the height of a building using the aerial photograph, the step comprising: Multi-level, correlating the texture of the building between the aerial photographs for an assumed height,
A three-dimensional map creation method, wherein a height of a building having a maximum correlation is determined as the height of the building.
成方法において、該３次元地図作成方法は該航空写真を
利用して建物の高さを求める工程を含み、該工程は建物
の高さを複数レベル仮定し、仮定された任意の高さにつ
いて該航空写真間の該建物のテキスチャの相関を求め、
相関が急激に変化する建物の高さを該建物の高さと決定
するようにしたことを特徴とする３次元地図作成方法。3. A three-dimensional map creation method using a plurality of aerial photographs, the three-dimensional map creation method including a step of obtaining a building height using the aerial photographs, the step comprising: Multi-level, correlating the texture of the building between the aerial photographs for an assumed height,
A method for creating a three-dimensional map, wherein the height of a building whose correlation changes rapidly is determined as the height of the building.
テキスチャの相関を求める前に、建物の形状を一致させ
る変換を行う工程を含むことを特徴とする請求項２及び
請求項３に記載の３次元地図作成方法。4. The method according to claim 2, wherein the step of determining the height of the building includes a step of performing a conversion for matching the shape of the building before determining the correlation of the texture of the building. 3D map creation method described.
求項１に記載の方法で建物の地上の位置座標を求め、求
めた建物の地上の位置座標による建物形状の方向及び向
きを一致させる変換を含むことを特徴とする請求項４に
記載の３次元地図作成方法。5. The transform for matching the shape of the building, wherein the position coordinates of the building on the ground are obtained by the method according to claim 1, and the direction and orientation of the building shape are determined based on the obtained position coordinates of the building on the ground. 5. The method according to claim 4, further comprising a conversion.
の屋根の形状を複数のパターンから選択し、選択された
屋根パターンについて相関を求める両者の航空写真で共
通に見える領域を求め、該領域についてテキスチャの相
関を求めるようにしたことを特徴とする請求項２〜請求
項５の何れか１に記載の３次元地図作成方法。6. The step of determining the height of the building includes selecting a shape of a roof of the building from a plurality of patterns, determining a correlation with respect to the selected roof pattern, and determining an area common to both aerial photographs. The three-dimensional map creation method according to any one of claims 2 to 5, wherein a texture correlation is obtained for the area.
錐形屋根、平屋根を含むことを特徴とする請求項６に記
載の３次元地図作成方法。7. The method according to claim 6, wherein the roof pattern includes a dome roof, a pyramid roof, and a flat roof.
面が傾斜又は平坦でない場合は、任意高さの水平面に２
次元地図を作製し、該水平面を基準として建物の高さを
求めるようにしたことを特徴とする請求項２〜請求項７
の何れか１に記載の３次元地図作成方法。8. In the method for creating a three-dimensional map, when the ground surface is not inclined or flat, a two-dimensional map is formed on a horizontal plane having an arbitrary height.
A two-dimensional map is prepared, and the height of the building is determined based on the horizontal plane.
The three-dimensional map creation method according to any one of the above.
元地図に含まれる建物の平面形状がポリゴンでない建物
は複数のポリゴンに分割して３次元地図を作製すること
を特徴とする請求項１〜請求項８の何れか１に記載の３
次元地図作成方法。9. The three-dimensional map creating method according to claim 1, wherein a building included in the two-dimensional map and whose planar shape is not a polygon is divided into a plurality of polygons to create a three-dimensional map. 9. The method according to claim 1, wherein
How to create a three-dimensional map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP26812199A JP2001091251A (en) | 1999-09-22 | 1999-09-22 | Three dimensional map preparing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP26812199A JP2001091251A (en) | 1999-09-22 | 1999-09-22 | Three dimensional map preparing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JP2001091251A true JP2001091251A (en) | 2001-04-06 |
Family
ID=17454189
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP26812199A Pending JP2001091251A (en) | 1999-09-22 | 1999-09-22 | Three dimensional map preparing method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP2001091251A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006185367A (en) * | 2004-12-28 | 2006-07-13 | Sharp Corp | Photovoltaic power generator installation support system and program |
| JP2007018130A (en) * | 2005-07-06 | 2007-01-25 | Nec System Technologies Ltd | Device for predicting amount of available solar energy, method for predicting amount of available solar energy, and program therefor |
| JP2008134648A (en) * | 2007-12-25 | 2008-06-12 | Geo Technical Laboratory Co Ltd | Method for producing 3-dimensional electronic map data |
| JP2009512048A (en) * | 2005-10-11 | 2009-03-19 | フランス テレコム | Method, apparatus and program for transmitting a roof and building structure in a three-dimensional representation of a building roof based on the roof and building structure |
| CN102706331A (en) * | 2012-06-21 | 2012-10-03 | 神翼航空器科技（天津）有限公司 | Correction method for aerial surveying and mapping images |
| CN105571573A (en) * | 2015-12-29 | 2016-05-11 | 南京中观软件技术有限公司 | Oblique photography position and attitude determining method and system thereof, and airplane position and attitude determining method and system thereof |
| CN105651264A (en) * | 2015-12-31 | 2016-06-08 | 中英海底系统有限公司 | Submarine cable detecting system |
| CN105651259A (en) * | 2015-12-29 | 2016-06-08 | 天津大学 | Five-degree-of-freedom aerial photograph measurement simulation system |
| CN106153012A (en) * | 2015-04-13 | 2016-11-23 | 苏州科沃斯商用机器人有限公司 | Specify spatial attitude measurement method of parameters and the application thereof of target |
-
1999
- 1999-09-22 JP JP26812199A patent/JP2001091251A/en active Pending
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006185367A (en) * | 2004-12-28 | 2006-07-13 | Sharp Corp | Photovoltaic power generator installation support system and program |
| JP4602074B2 (en) * | 2004-12-28 | 2010-12-22 | シャープ株式会社 | Photovoltaic generator installation support system and program |
| JP2007018130A (en) * | 2005-07-06 | 2007-01-25 | Nec System Technologies Ltd | Device for predicting amount of available solar energy, method for predicting amount of available solar energy, and program therefor |
| JP4718918B2 (en) * | 2005-07-06 | 2011-07-06 | Ｎｅｃシステムテクノロジー株式会社 | Usable solar energy amount prediction apparatus, usable solar energy amount prediction method and program |
| JP2009512048A (en) * | 2005-10-11 | 2009-03-19 | フランス テレコム | Method, apparatus and program for transmitting a roof and building structure in a three-dimensional representation of a building roof based on the roof and building structure |
| JP2008134648A (en) * | 2007-12-25 | 2008-06-12 | Geo Technical Laboratory Co Ltd | Method for producing 3-dimensional electronic map data |
| CN102706331A (en) * | 2012-06-21 | 2012-10-03 | 神翼航空器科技（天津）有限公司 | Correction method for aerial surveying and mapping images |
| CN102706331B (en) * | 2012-06-21 | 2014-01-01 | 神翼航空器科技（天津）有限公司 | Correction method for aerial surveying and mapping images |
| CN106153012A (en) * | 2015-04-13 | 2016-11-23 | 苏州科沃斯商用机器人有限公司 | Specify spatial attitude measurement method of parameters and the application thereof of target |
| CN105571573A (en) * | 2015-12-29 | 2016-05-11 | 南京中观软件技术有限公司 | Oblique photography position and attitude determining method and system thereof, and airplane position and attitude determining method and system thereof |
| CN105651259A (en) * | 2015-12-29 | 2016-06-08 | 天津大学 | Five-degree-of-freedom aerial photograph measurement simulation system |
| CN105651264A (en) * | 2015-12-31 | 2016-06-08 | 中英海底系统有限公司 | Submarine cable detecting system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA3103844C (en) | Method for reconstructing three-dimensional space scene based on photographing | |
| US20240153143A1 (en) | Multi view camera registration | |
| CN103810744B (en) | It is backfilled a little in cloud | |
| JP4245963B2 (en) | Method and system for calibrating multiple cameras using a calibration object | |
| US9210404B2 (en) | Calibration and registration of camera arrays using a single circular grid optical target | |
| JP4680796B2 (en) | Image base protrusion displacement mapping method and double displacement mapping method using the method | |
| CN108289208A (en) | A kind of projected picture auto-correction method and device | |
| CN105026997A (en) | Projection system, semiconductor integrated circuit, and image correction method | |
| JP6174968B2 (en) | Imaging simulation device | |
| JP2015219679A (en) | Image processing system, information processing device, and program | |
| CN111161394B (en) | Method and device for placing three-dimensional building model | |
| JP2002310619A (en) | Measurement device | |
| JP3690581B2 (en) | POSITION DETECTION DEVICE AND METHOD THEREFOR, PLAIN POSITION DETECTION DEVICE AND METHOD THEREOF | |
| JP2001091251A (en) | Three dimensional map preparing method | |
| Draréni et al. | Methods for geometrical video projector calibration | |
| CN112912936A (en) | Mixed reality system, program, mobile terminal device, and method | |
| CN106023723A (en) | Educational toy set and convex mirror imaging correction method | |
| WO2022078438A1 (en) | Indoor 3d information acquisition device | |
| JP3791186B2 (en) | Landscape modeling device | |
| JP2000329552A (en) | Three-dimensional map preparing method | |
| US20210224538A1 (en) | Method for producing augmented reality image | |
| JP2008052403A (en) | Pattern with coded two-dimensional location information and location identification system and method using the pattern | |
| CN111179347B (en) | Positioning method, positioning equipment and storage medium based on regional characteristics | |
| Lai et al. | Projective reconstruction of building shape from silhouette images acquired from uncalibrated cameras | |
| Jones | Augmenting complex surfaces with projector-camera systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20060919 |
|
| A711 | Notification of change in applicant |
Free format text: JAPANESE INTERMEDIATE CODE: A711 Effective date: 20060922 |
|
| A131 | Notification of reasons for refusal |
Effective date: 20080812 Free format text: JAPANESE INTERMEDIATE CODE: A131 |
|
| A02 | Decision of refusal |
Free format text: JAPANESE INTERMEDIATE CODE: A02 Effective date: 20081209 |