JPH02253378A - Automatic deciding device for photographing position - Google Patents

Abstract

PURPOSE:To secure a fixed difference of information between pictures by preparing a key position extracting part and selecting a photographing position based on the form information on an object and the pattern information of the object surface so as to secure an approximately fixed changing value of the extracted surface information between the pictures contiguous to each other. CONSTITUTION:The form information inputted via a form information input part 1 is approximately divided into plural polygonal planes omegai and at the same time the correspondence is secured between the planes omegai and the pattern information inputted via a pattern information input part 2 via a form/pattern information approximation part 3. A surface information changing value detecting part 4 detects the changing value of the surface information based on the changes of visual points to the planes omegai. Then a key position extracting part 5 extracts a key longitude and a key latitude of a photographing position so that the changing value of the surface information detected by the part 4 is approximately fixed between pictures contiguous to each other. Then a key photographing position 6 is set at the key longitude and key latitude extracted by the part 5. As a result, a fixed difference of information is kept between pictures.

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical field to which the invention pertains) The present invention relates to a device that automatically determines the photographing position when photographing real objects such as products, works of art, animals and plants from a plurality of directions. Image information photographed from a photographing position determined according to the present invention is stored in a center device, and is used by an image information providing device or the like that searches and provides information to a terminal device via a transmission path.

(Prior art) When photographing real objects such as products, works of art, animals and plants from multiple directions, a conventional method for determining the photographing position is to use a virtual hemisphere that covers the object O as shown in Figure 8. O
There is a method of assuming A and determining photographing positions at equal angular intervals in the longitude direction ψ (horizontal direction) and latitude direction λ (vertical direction) of this virtual hemisphere.

However, in this case, as the latitude λ of the virtual hemisphere ○A increases, the density of photographing positions increases as it approaches the pole OC, and the image photographed from near the pole (although the attitude of the object within the screen changes) is substantially They have almost the same information. Therefore, it can be seen that it is necessary to keep the distance between the shooting positions on the hemisphere constant.

However, even if multiple shooting positions are determined in this way, the difference in information that the viewer receives from images taken from each shooting position will vary greatly depending on the shape of the object and the shooting position. .

For example, compare the image of a rectangular parallelepiped shown in Figure 9 viewed from the direction of ■ (Figure 1θ (a)) and the image viewed by rotating it by 10 degrees (Figure 10 (b)). The differences between the images when viewed from the direction of
1 (b)), the difference when viewed from the direction of ■ is clearly felt to be larger.

This means that even if the interval between the shooting positions on the hemisphere is constant, the difference in information between each image will not be constant. Therefore, when these images are displayed one after another in the order of adjacent imaging positions, the information received by the viewer becomes larger in one section and smaller in another section. In other words, when such a set of images is stored, the storage efficiency in terms of how much information is actually given to the viewer deteriorates.

Conventionally, in order to solve such problems, a method has been proposed for automatically determining a photographing position where the difference in information between images becomes constant. In this conventional method, the sum of the amount of information generated and the amount of information lost as the viewpoint position moves is calculated based only on the shape of the object to be photographed, and is used as the difference in information between each image.

However, in reality, pattern information such as color exists on the surface of the object to be photographed. Therefore, the difference in information between each image should include changes in pattern information, but this was not included in the conventional method.

For this reason, even if the information difference between images depending on the shape is small, the information difference between each image will not be constant if the information varies greatly depending on the pattern.

(Objective of the Invention) An object of the present invention is to identify the point where the difference in information between the captured images occurs when a plurality of images are captured along a virtual hemisphere covering the object. An object of the present invention is to provide a device that automatically determines a photographing position at which the difference in information between images becomes constant by solving the problem based on pattern information on the surface.

(Structure of the Invention) (Characteristics of the Invention and Differences from the Prior Art) The present invention provides information on the information between each image when capturing a plurality of images along a virtual hemisphere covering the object O. The most important feature is that the amount of surface information change corresponding to the difference is determined based on the object shape and the pattern on the object surface, and the photographing position is determined so that this amount is constant between each image.

In the present invention, the amount of surface information change based on the object shape and the pattern on the object surface is defined as follows. First, the 31st ff.
The object O to be photographed shown in l(a) (hatching means a pattern) is formed into fine polygons ω1 (i=o) such as m triangular patches as shown in FIG. 3(b).

1, . . . 1-1) for plane approximation. It is assumed that each polygon ωi is given color information at a corresponding position of the object O.

As shown in Figure 4, first place! When an object O is photographed from Iu, if the object O is photographed from a 9th position V, where su is the area where a certain polygon ωi is projected onto the screen surface, then the projection of the polygon ωi onto the screen surface is It is assumed that the projection area is sv.

Let n types of colors on the surface of object 0 be Ij (j=o,...n
-1), and if the projected area of n types of colors Ij projected onto the area su is pj, then the projected area of each color is p(s u) = (po, , , -,
pn-1) (1) Vector p(su)
It can be expressed as Assuming parallel projection for simplicity, place 1iu and place l! The change di in the projected area of each color of the polygon ωi projected on the screen surface between the two images taken from ■ is di −p(su) p(sv) = (Δpop−−−gΔp n−1) (2)
If the component Δpj of di is positive, it can be considered that the information about the n types of colors Ij has increased by an amount proportional to Δpj due to the movement of the viewpoint.

If it is negative, it can be considered that the information has decreased by an amount proportional to -8pj.

Therefore, the norm of the change di in projected area obtained for each polygon ωi [d i l summation Duv
can be regarded as the amount of change in information between an image taken from a position [u] and an image taken from a position V of the object O.

Duv will be referred to as surface information variation amount.

In the present invention, we focus on the fact that the amount of change in surface information is proportional to the amount of change in information between two images, and that it can be extracted only from the shape information of the object and the pattern information on the surface. The feature is that the difference in information between each image is made constant by selecting the shooting position so that it is almost constant between the images.

On the other hand, in the conventional technique, the photographing positions are simply determined at equal intervals, so the amount of change in information regarding the object ○ between each image is not constant.

(Embodiment) FIG. 1 is a block diagram of the basic configuration of the present invention, and in the figure, 1 is a shape information input unit for inputting shape information of an object;
2 is a pattern information input unit for inputting pattern information on the surface of the object;
3 approximately divides the shape information input from the shape information input section 1 into a plurality of polygon planes ωi, and associates the pattern information input from the pattern information input section 2 with the plurality of polygon planes ωi. Shape/pattern information approximation part, 4
5 is a surface information change amount detection unit that detects a surface information change amount based on a viewpoint change with respect to a plurality of polygonal planes ωi that are plane-approximated by the shape/pattern information approximation unit 3 and are associated with pattern information; A representative position extractor 6 extracts representative longitudes and representative latitudes of photographing positions so that the surface information change amount detected by the surface information change amount detector 4 is approximately constant between adjacent images; 6 is a representative position extractor; The representative photographing position set to the representative longitude and representative latitude extracted in section 5 is shown.

The above-described specific configuration is shown in FIG. 2, and the shape/pattern information approximation section 3 includes a shape approximation section 31, a pattern approximation section 32, and a shape/pattern information storage section 33. The surface information change amount detection section 4 and the latitude direction change amount measurement section 41.

It consists of an averaging section 42 and a longitudinal change measurement section 43,
The averaging section 42 further includes an accumulation section 421 and a calculation section 422.
It becomes. The representative position extraction unit 5 includes a representative latitude extraction unit 51 and a representative longitude extraction unit 52, and both of these units include a storage unit 511.521 and an extraction unit 512.522, respectively.

In addition, the representative shooting position 6 has a representative latitude 61 and a representative longitude 62.
are output from each extractor 512 and 522 of the representative position extractor 5, respectively.

To operate this, the shape information of the object is input to the shape information input section l. Further, pattern information (color information) on the surface of the object is input to the pattern information input section 2. The shape approximation unit 31 of the shape/pattern information approximation unit 3 reads the shape information of the object from the shape information input unit 1, and converts the shape of the object into m minute polygons ωi(i”
otl*...m-1), and the shape information of each polygon ωi and its outward normal vector information are stored in the shape/[-like information storage unit 33.

On the other hand, the pattern approximation unit 32 reads pattern information (color information) on the surface of the object from the pattern information input unit 2, and calculates n types of color schemes j (
j:0. ..., n-1), and each polygon ω
It is stored in the shape/pattern information storage section 33 in association with i.

The latitudinal change amount measurement unit 41 of the surface information change amount detection unit 4 reads the shape information of the object plane-approximated by polygons and the pattern information associated with each polygon ωi from the shape/pattern information storage unit 33. , based on this, as shown in Fig. 5, the amount of surface information change D when the longitude of the photographing position of the object is fixed at ψ = 0 and the latitude λ is moved by minute angles Δλ.
Obtain O(λ) from equation (3). However, the angle Δλ=π/
(2・C0) (CO is a natural number).

Next, move the longitude ψ by a minute angle Δψ and do the same to get D
Find Δψ(λ). However, Δφ = 2π/C1 (C1 is a natural number). This is repeated until the latitude ψ = 2π, and the amount of change in surface information in the latitude direction Dφ (λ) (φ = Δψ,
2Δψ# ++1+l 2π, λ=Δλ.

2Δλp mamg π/2), and the averaging unit 42
The data is stored in the storage unit 421 of.

The calculation unit 422 reads out the surface information change amount Dψ(λ) in the latitude direction accumulated in the storage unit 421, and calculates the average value Da(λ) for the longitude ψ (λ=Δλ, 2Δλ@ass@π/2)
(4) is calculated, and the storage unit 511 of the representative latitude extraction unit 51
Accumulate in.

The extraction unit 512 extracts the average value Da(λ)(λ
=Δλ, 2Δλe −−−e f/2), and using the predetermined interval value δ of the amount of change in surface information in the latitude direction, the average value Da is calculated as shown in FIG.
Divide the integral value of (λ) equally. Then, the X latitudes Ai (i=0., . . . , x-1) serving as the division positions are outputted as representative latitudes 61, and are also outputted to the longitudinal direction variation measuring section 43.

The longitudinal change measurement unit 43 is the shape/pattern information storage unit 33
The shape information and pattern information obtained by plane approximation of the object shape are read out from , and then the latitude A 1 (i=o, , , , x-1
), and in the same way as the latitudinal direction variation measuring unit 41, the surface information variation DAi(ψ) (C
o.

***gX−12ψ=Δψ, 2Δψe**mg2π) is measured and stored in the storage unit 521 of the representative longitude extraction unit 52.

Surface information change amount D in the longitudinal direction accumulated in the accumulation section 521
Ai(ψ) is read out, and in the same manner as the first representative latitude extraction unit 51, the integrated value of DAi(ψ) is equally divided by the interval value δ as shown in FIG. Then, yi longitude marks lj (j”Ot*e-s y-1e l”Ot-
, , x-1) and outputs it as the representative longitude 62.

Although this embodiment has been described on the assumption that the interval value δ of the amount of change in surface information is preset inside the apparatus, this interval value δ may be input from the outside.

If the representative latitude and longitude are determined based on the shape information and pattern information of the object in this way, the image taken of the object from the shooting position on the virtual hemisphere determined by the representative latitude and representative longitude will be the same as the adjacent image. The amount of change in surface information between the two becomes almost constant.

(Effects of the Invention) As explained above, the amount of change in surface information in the present invention is extracted based on shape information and pattern information. It is possible to keep the information difference between images constant. Therefore, when these images are displayed continuously, they can be displayed more smoothly than when objects are photographed at equal intervals. Furthermore, when storing these images in a database or the like, it is possible to improve the storage efficiency in terms of the amount of information actually given to the viewer, compared to simply taking images at equal intervals.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the basic configuration of the present invention. Fig. 2 is a block diagram showing its specific configuration, Fig. 3 is an explanatory diagram of plane approximation by polygon ωi of the object to be photographed,
Fig. 4 is a diagram showing the projected area pj on the screen surface when polygon ωi in Fig. 3 is photographed from positions u and v, Fig. 5 is an explanatory diagram for calculating the amount of change in surface information of an object, and Fig. 6 is an explanatory diagram when the integral value of the average value Da(λ) is equally divided using the predetermined interval value δ of the amount of surface information change in the latitude direction, and FIG. An explanatory diagram when the integral value of the average value DAi (ψ) is equally divided using the interval value δ of the amount of change in surface information, and FIG. 8 shows the determination of the photographing position of the object in the longitude and latitude directions using the conventional method. The explanatory diagram, FIG. 9, is a diagram of the rectangular parallelepiped viewed from two directions, and FIGS. 10 and 11 are diagrams showing differences in images when the rectangular parallelepiped of FIG. 9 is viewed from two directions. 1... Shape information input section, 2... Pattern information input section, 3... Shape/pattern information approximation section, 4... Surface information change amount detection section, 5... Representative position extraction section, 6 ...
Representative photographing position, 31... Shape approximation section, 32... Pattern approximation section, 33... Shape/pattern information storage section, "41...
・Latitude direction change measurement unit, 42... Averaging unit, 43・
. . . Longitudinal direction change amount measuring section, 51 . . . Representative latitude extraction section, 52 . . . Representative longitude extraction section. 61... Representative latitude, 62... Representative longitude, 421
゜511, 521...Storage unit, 422...Calculation unit,
512, 522...extraction section. (a) (b) (a) (b) (a) (b) (b) (a) (b) (b) (a) (b) (b) (b) (a) (b) (b) (b) ψ longitude diagram (b) (b)

Claims (1)

[Claims] a shape information input section for inputting shape information of an object; a pattern information input section for inputting pattern information on the surface of the object; and approximately dividing the shape information input from the shape information input section into a plurality of polygonal planes. , and a shape/pattern information approximation unit that associates the pattern information input from the pattern information input unit with the plurality of polygonal planes; A surface information change amount detection section that detects a surface information change amount based on a viewpoint change for a plurality of attached polygonal planes; An automatic photographing position determination device comprising: a representative position extracting unit that extracts a representative longitude and latitude of a photographing position so that the representative longitude and latitude of the photographing position are constant.