JPH03269680A - Method for processing three-dimensional display - Google Patents

Abstract

PURPOSE:To attain executing display with depth perception to a picture by giving pseudo color correction based on color aberration corresponding to distance in a depth direction. CONSTITUTION:Various aspects are fetched by a camera input part 1 having a two-lens camera and the distribution of the parallax quantity of respective objects within fetched images is found by a parallax quantity detecting and processing part 2. Pseudo color conversion is given to a picture inputted in a pseudo color converting and processing part 3 corresponding to the obtained parallax quantity. Picture information fetched by plural cameras at a three- dimensional display part 4 is reproduced and is displayed through a lenticular lens. That is, the parallax quantity of the respective objects is calculated at the parallax quantity detecting and processing part 2 to the picture fetched by the two-lens camera of the camera input part 1. Then, depth perception is made increased by giving pseudo color correction corresponding to a distance in a depth direction at the pseudo color converting and processing part 3.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a stereoscopic display processing method in the field of color image processing that enhances depth perception by utilizing a specific combination of hues. .

[Conventional technology]

One of the technical problems is faithfully reproducing colors in image information input by a camera.

For example, it is necessary to measure the spectral luminance characteristics of a TV camera for input, a color image display for image output, etc., and perform digital processing to correct them so that they are equal. On the other hand, there are displays that artificially emphasize or change colors. FIG. 12 is a diagram illustrating a false color composite display of color images from two artificial satellites. The image information captured from the satellite is received by the earth station.
Is displayed. Observation wavelength band of color band from artificial satellite and R of image display device of color image display
, G, and B are different from each other.

Band wavelength 101 of the image information from the satellite where the hue of the reproduced color image is far from the natural one
The false color composite display 102 is performed by reversing the order of the images, and the green vegetation 103 near Japan that has been transmitted to the earth is displayed in red. As a result, the contrast with the area 104 with a small distribution of green vegetation becomes clear, and analysis from one image becomes easy.

In this way, the two false color combination display is effective as an auxiliary means when one person can read the image.

FIG. 13 is a diagram illustrating pseudo-color display for a black-and-white X-ray image. Since the X-ray image 110 is a single-band grayscale image, considerable skill is required to find a lesion in the lungs, for example. In addition, if two lesions overlap each other, determining whether one of them is closer to the other in the depth direction is very important in determining the opening if surgery is required. . That is, a scale in which a specific color is assigned to each level of gradation is prepared in advance, and each pixel of a gradation image to be displayed in two is displayed in a corresponding color using this color scale. Using the X-ray image 112 subjected to such pseudo-color conversion, objects that were difficult to recognize in the original black-and-white gradation image can now be easily recognized, contributing to the discovery of lesions and the like. However, there remains a problem in understanding the relative positions of various lesions in the depth direction of the M image.

[Problem to be solved by the invention]

Conventional pseudo color conversion techniques are aimed at improving the VL recognition of objects in two-dimensional image information, and no pseudo color conversion technique for three-dimensional image information has yet been found.

The present invention solves the problems of the conventional method and performs pseudo color correction based on chromatic aberration depending on the distance in the depth direction.
The purpose is to enhance the depth perception effect for images.

[Means to solve the problem]

FIG. 1 shows a basic configuration for implementing the method of the present invention. Various quantities are taken in from a camera input unit 1 having a 2@ camera, and a parallax amount detection processing unit 2 calculates the distribution of the amount of parallax for each object in the taken video. The pseudo color conversion processing unit 3 performs pseudo color conversion on the input image according to the obtained amount of parallax. Image information captured by a plurality of cameras is reproduced and displayed on the stereoscopic display unit 4 via a lenticular lens.

[For production]

In the present invention, the parallax amount detection processing section 2 calculates the amount of parallax of each object for the image captured by the twin-lens camera of the camera input section 1, and performs pseudo-color correction according to the distance in the depth direction. is performed by the pseudo color conversion processing section 3 to increase depth perception.

In the prior art, although there are examples of pseudo-color conversion applied to two-dimensional images, there are no examples of pseudo-color conversion applied to three-dimensional images to increase depth perception.

[Example] FIG. 2 is an explanatory diagram regarding a camera manual section and a stereoscopic display section. Various quantities [11 are taken in from a plurality of cameras 12, and the information from the plurality of cameras is separated and reproduced using a lenticular lens 13. Fine structure 14 of lenticular lens 13
In this method, scanning lines are arranged periodically according to the number of cameras per pitch of the lens.

FIG. 3 is a diagram for explaining the relative positions of 93 solids. Cube 219 Cylinder 22. The relative positions of the triangular pyramids 23 are arranged in the order of 29, 28, and 27 from the camera side. Also, when viewed from the triangular male side, it looks like images 24, 25, and 26.

FIG. 4 shows the principle for obtaining the depth in the depth direction when 2@cameras are used. Three solids 31, 32.33
is captured from the twin-lens camera 34. As a result, the solid 32.
The congestion angles 3736 and 35 increase in the order of 31.33. The images input to each camera are shown in 38.39, and there is a deviation in the relative positions of the three solids of the image from the left camera and the image from the right camera. The relative pressure M 40 between the cylinder 31 and the triangular pyramid 33 in the Wi image 38 is the image 39
On the other hand, the relative distance 42 in Wi image 38
The relative distance 41 between the triangular pyramid 33 and the cube 32 is 1 image 3
The relative pressure M at 9 is shorter than that of 43. The difference in relative distance between the objects in these two images is determined as the amount of parallax.

In general, one person's depth perception is 600 for each eye, and objects 601 and 60 whose depths differ by ΔD (604)
It depends on how well you can distinguish between 2 and 2. That is, it is defined as ΔD/D (605).

FIG. 5 is an explanatory diagram regarding the relationship between two hues and depth perception. As shown in FIG. 5(B), the coordinates of the RGB color space are transformed to form an H (hue), S (saturation), and (brightness) space. Using the H3V space is effective when considering color components that cannot be separated using the RGB space. This H
In the 3V model, the origin of the brightness axis is represented by black BL,
The maximum value is represented by white W, and the saturation S of each hue is proportional to the brightness V. Figure 5 (A) shows one hue divided into six equal parts (red, yellow, green, cyan) based on the red hue.

The results of a subjective evaluation experiment of the distribution of depth perception in each hue (blue, magenta) are shown.

The experiment was conducted in such a way that the subject was given a three-dimensional display using a single liquid crystal shutter, as shown in FIG.

A sine wave pattern as shown in 706 is generated by the chromaticity signal pattern generator 700, and the left eye signal 702. is shifted in phase by 1/60th of a second. A synchronization signal 704 is sent from the 0 chromaticity signal pattern generator 700 that sent the right eye signal 703 to the stereoscopic display section 701 to the liquid crystal shutter type glasses 705.
The left and right eyes are configured to receive a reproduced image independently. The amplitude 707 of the sine wave corresponds to the depth in the depth direction. We conducted an experiment by placing each hue on this sine wave.

The vertical axis of the graph shown in FIG. 5(A) represents the depth sensitivity obtained from the subject, and the horizontal thin lines represent the hue.

As is clear from the figure, the depth sensitivity is quite different for each hue. 0 Hue 2 (near green)
The depth sensitivity was the worst and the best results were obtained with 0 hue 4 (near blue).Over 0 To summarize, 1 hue 2, 1, 0, 3,
It was shown that the depth sensitivity changes in the order of 5.4. Based on this data, we created two types of two-dimensional plan views with various combinations of hues, and conducted subjective evaluation experiments in the same manner as above. The first hue combination is shown in Figure 5 (C
) As shown in the diagram, the hue is 2.1° 0.3.5° from the center of the circle.
Chart 43 is colored in the order of 4. As a result, 80% of the subjects perceived the deepest part to be in the center of the circle compared to the outside of the circle. Similarly, as shown in FIG. 5(D), the hue (from the center of the circle, 4
Chart 44 was created by combining 5.30.1.2). As a result, 75% of the subjects perceived a triangular pyramid solid with the vertex in the foreground.

We also conducted a subjective evaluation experiment by displaying concentric planar figures on a display. Each day is 1 hue O.

Colors corresponding to 60°, 120°, 180°, 240°, and 300" were arbitrarily combined.The brightness and saturation were kept constant.As a result, for example, green → magenta → blue from the center side. By changing the order of the combinations, such as blue → magenta → green from the middle side, a difference occurred in the depth perception obtained by monocular observation.

Yellow → magenta → cyan from the center side,
In the case of cyan → magenta → yellow from the middle side, only an emphasis effect was produced due to visual characteristics.
When hues are combined concentrically at regular intervals, a kind of optical illusion is thought to occur due to visual characteristics.

FIG. 7 is an explanatory diagram illustrating an example in which pseudo color correction is applied to an input monochrome stereoscopic image according to the amount of parallax. Reference numeral 51 shown in FIG. 7(A) is a monochrome stereoscopic image, and a correlation diagram 55 between color gradation and the amount of parallax is created as shown in FIG. 7(B) according to the amount of parallax. The monochrome stereoscopic image 51 still has the problem of poor sense of distance between objects.

Therefore, pseudo colors are added to each of the 52, 53, and 54 objects according to the amount of parallax. As a result, the depth perception was enhanced as shown in Figure 7 (C).

FIG. 8 is an example in which hue conversion is applied to the sides of each object in the depth direction. Input monochrome 3D image 51
The difference from the case shown in FIG. 7 is that, in accordance with the amount of parallax, one hue conversion is applied to the original image as in the case shown in FIG. 7, and hue conversion is also applied to each side. As a result, the depth perception became even stronger than in the case shown in FIG.

FIG. 9 is an explanatory diagram of X-ray images of the chests of nine people. The X-ray image (the slight IA light in &61 is basically subjected to pseudo-color conversion.

However, the relative positions of ribs, spine, etc. in the depth direction are not necessarily sufficient from the gray scale image.Therefore, based on a certain amount of gray scale image information, the depth is linearly converted as shown in Figure 9(B). After that, as shown in FIG. 9(C), a hue is applied according to the degree of depth. FIG. 9(D) shows an example of an image obtained by applying pseudo-color conversion to the original X-ray image, and shows the structure 65 in the depth direction near the chest of person 0.
Since 66°67, 68, 69, 70, and 71 are relatively emphasized by 1 due to the difference in 9 hues, it is convenient for searching for disqualified parts near copy 111.

FIG. 10 is an explanatory diagram of a capillary image taken with the a-microscope. Conventionally, analyzing an image of a blood vessel involves analyzing an image 82 in which a three-dimensional capillary structure is projected onto a two-dimensional plane.

Using the slight :a difference information as a clue, the main blood vessels and branch blood vessels are processed separately. However, based on the grayscale image shown in FIG. 10(A), FIG. 10(B)
After linearly applying the depth as shown in the figure, nine hues are applied according to the depth. As a result, Figure 10 (C
) As shown in the figure, it is possible to clearly understand the relative distances of capillaries that overlap each other depending on the depth,
convenient.

〔Effect of the invention〕

As explained above, according to the present invention, compared to the conventional full-color stereoscopic display, as shown in the results of the subjective evaluation experiment shown in FIG. Comparing Method B (a method in which motion parallax characteristics are introduced into binocular parallax characteristics) and Method C using the present invention, depth perception has become even stronger. In addition, the feeling of fatigue when observing for a long time was significantly reduced, and a reconstructed image of the natural landscape was obtained.

[Brief explanation of drawings]

Fig. 1 is a flowchart of the process in the case of the present invention that performs pseudo color conversion according to the amount of parallax, Fig. 2 is a diagram explaining the camera input section and the stereoscopic display section, and Fig. 3 is a relative view of the input scenery. Positional relationship diagram, Figure 4 is a diagram explaining the principle of detecting the amount of parallax from a twin-lens camera, Figure 5 is an explanatory diagram regarding hue and depth perception, Figure 6 is a diagram of the experimental system, and Figure 7 is a diagram explaining the amount of parallax. Figure 8 is an explanatory diagram illustrating an example in which hues are assigned according to the amount of parallax. Figure 8 is an explanatory diagram illustrating an example in which hue conversion is applied to each object according to the amount of parallax. Figure 9 is an image in the depth direction for an X-ray image. FIG. 10 is an explanatory diagram illustrating an example of correction; FIG. 10 is a diagram illustrating a case in which hue conversion is applied to a capillary image; FIG. 11 is a diagram illustrating depth perception between the present invention and the conventional method;
FIG. 13 is a diagram for explaining a false color composite display of color images from an artificial satellite, and FIG. 13 is a diagram for explaining a false color display for a monochrome X-ray image. In the figure, 1 represents a camera input section, 2 represents a parallax amount detection processing section, 3 represents a pseudo color conversion processing section, and 4 represents a stereoscopic display section.

Claims (3)

[Claims] (1) For an image containing a subject whose distance in the depth direction is known, a display with stereoscopic perception can be performed by creating and displaying an image in which the subject is colored with a specific hue according to the distance in the depth direction. Characteristic three-dimensional display processing method. (2) A stereoscopic display processing method characterized by adding a specific hue according to the amount of parallax to a stereoscopic image using the visual characteristics of binocular parallax to display a display with enhanced depth perception. (3) A stereoscopic display processing method characterized in that when displaying a monochrome stereoscopic image, depth perception is enhanced by using a specific combination of hues depending on the depth.