JP2002230558A - Method of evaluating visual environment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method of evaluating a visual environment capable of evaluating a visual environment in a manner close to recognition by a human viewing field. SOLUTION: The method includes a digital image introduction means and a digital image analysis means. Images introduced by the introduction means are subjected to contrast measurements and frequency component analysis by the analysis means so as to determine the condition of a visual environment while taking ambient optical conditions such as ambient brightness into account to allow evaluation of the visual environment in a manner close to human recognition. The use of digital images allows significant simplification in terms of devices and facilitates processes.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for evaluating changes in the visual environment mainly due to heavy fog, snowstorm, etc., which lower visibility on a road.

[0002]

2. Description of the Related Art For example, on a road on which a vehicle passes, changes in the state of the visual environment due to dense fog or snowstorm are evaluated.
Recognizing the driver in advance by a sign or the like is a very useful safe driving support means. Particularly in cold regions, coastal areas with large differences in temperature, and mountainous regions, there is a phenomenon in which the field of view is suddenly closed by snowstorm or dense fog, and a change in the visual environment on the road sometimes even affects human life.

Conventionally, when judging a change in visibility, an object having a light-colored portion and a dark-colored portion is observed from a fixed point at arbitrary time intervals, and the brightness of the light-colored portion and the dark-colored portion are measured.
The visual environment was evaluated by measuring the luminance difference, that is, the contrast. When the visibility is good and the object whose contrast is to be measured is clearly visible, such as on a sunny day, the contrast is large. In the case of dense fog or snowstorm, the object whose contrast is to be measured is blurred and difficult to see, and the contrast is reduced.

[0004]

However, the visibility of a visual object is determined by three factors: contrast, background luminance, and size of the visual object, but the visual object is large or the surroundings are bright. If the contrast is low, the visibility is increased even if the contrast is low. Conversely, if the luminous body is low in visibility at night and has a low luminous intensity, a large contrast can be obtained. Therefore, the method of evaluating the visual environment only by the contrast does not always match the visibility of the human visual field.

[0005] Further, the contrast is usually measured in a limited narrow range, which is different from the field of view of an actual driver or the like, and lacks perspective, left and right, and means for observing the visual environment, contrast, and the like. In some cases, the correlation between the visual environment and the contrast may not be established due to the object to be measured being shaded, wet, or contaminated.

Accordingly, the present invention provides a visual environment evaluation method capable of evaluating a visual environment in a form close to recognition in the human visual field.

[0007]

Means for Solving the Problems In order to solve the above-mentioned problems, the present inventors have conducted intensive studies and as a result, as described above, it is useful to measure contrast as described above. The visual environment can be evaluated in a manner closer to human recognition by taking into account the surrounding optical conditions such as the brightness of the image. Numerical analysis, the contrast between the virtual object and the virtual background set on the image, and decomposition of the image into frequency components
The present inventors have learned that the image can be easily obtained by characterizing an image based on the distribution of low frequency components, and have completed the present invention.

[0008] That is, a visual environment evaluation method according to the present invention includes a digital image introducing means and a digital image analyzing means, wherein an image introduced by the digital image introducing means is subjected to contrast measurement and frequency conversion by the digital image analyzing means. The state of the visual environment is determined by analyzing the components.

According to the present invention, in addition to the measurement of the contrast, the surrounding optical conditions such as the brightness of the surroundings are measured by analyzing the frequency components. The environment can be evaluated. Since the measurement of the contrast and the measurement of the surrounding optical conditions are performed using digital images, the measurement can be extremely simplified and easily performed.

Further, since the measurement of the surrounding optical conditions is performed by analyzing the frequency components of the image, it is possible to analyze not only a part of the image but also a wide range. By using together, it is possible to evaluate the correlation with the visual environment to be visually recognized with higher accuracy and higher. In addition, since the evaluation can be performed over a wide range of the image, even if the target object changes due to contamination or the like, the evaluation can be performed as a whole, so that the influence of the change in the state of the target object on the evaluation can be reduced.

The contrast measurement may be a method of outputting a relative degree with respect to a previously obtained limit contrast. The limit contrast is when the brightness of the visual object gradually decreases from a value larger than the background brightness while keeping the background luminance and the size of the visual object constant, and the visual object becomes invisible The luminance is obtained, the luminance difference between the luminance at that time and the background luminance, that is, the contrast, is referred to, and the relative evaluation is performed by comparing the limit contrast with the contrast value obtained by the contrast measurement. It is more similar to the information obtained by human vision than the value.

The analysis of the frequency components may be a method of converting a digital image into frequency components by a two-dimensional orthogonal transform fast Fourier transform, and analyzing the distribution of high and low frequencies.

The analysis of the frequency components may be performed by obtaining a power spectrum for each digital image unit and calculating an average of the power spectra within a predetermined area. Obtaining the average of these within a predetermined area enables evaluation of the visual environment over a wide range.

The image introduced by the digital image introduction means may be divided into a plurality of areas, and the contrast measurement and the analysis of the frequency component may be performed for each area by the digital image analysis means. By dividing a digital image into a plurality of regions such as perspective, right and left, and performing contrast measurement and frequency component analysis on each region, evaluation in a form closer to human vision can be performed.

The light emission luminance of the self-luminous body installed on the road or in the vicinity thereof may be changed based on the visual environment state obtained by the invention according to claim 1, 2, 3, 4 or 5. . In the case of dense fog or snowstorm, the visibility can be improved by increasing the emission brightness from the normal case, and by changing the emission brightness according to the situation, the emission brightness of the self-luminous body is always appropriate. Become.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The embodiments of the present invention will be specifically described below. FIG. 1 is a diagram showing a flow of a method for evaluating a visual environment. The object and its surroundings are converted into digital images by the digital image introducing means. The digital image introduction means may be any device capable of digitizing an image, such as an ITV camera installed on a road, a vehicle-mounted camera in a road patrol car, a digital camera installed at a fixed point, a handy camera shot on site, and the like. The digital image is transmitted to a digital image analysis unit, where a contrast measurement and a frequency analysis are performed.

In the digital image analysis means, the contrast is measured by comparing the object set in the image with the background part. The RGB values of the image in the object and the background are converted into voxel values, the average voxel value in the extraction range is calculated, and the luminance difference is calculated to obtain the contrast. When the visibility is reduced due to dense fog or snowstorm, the brightness difference between the object and the background is reduced, and the contrast is reduced. The limit contrast is determined from the approximate size of the target object and the distance from the target object, and the obtained contrast is compared with the limit contrast to obtain a VISIBILITY LEBEL that is more similar to human vision.
(Hereinafter abbreviated as VL).

For the analysis of frequency components, digital images usually have a brightness component from white to black, ie from 0 to 255.
Are represented by 256 gradations, but the brightness density distribution of each image unit constituting the digital image can be quantified as a frequency component. The image is decomposed into frequency components, and the image is characterized by the distribution of high frequency and low frequency components. For example, a relatively bright case has a high frequency, and a relatively dark case has a low frequency. Further, when the visibility is good, a complicated frequency distribution is obtained, and when the visibility such as dense fog or snowstorm deteriorates, the frequency distribution becomes simple. To digitize the frequency component, for example, a two-dimensional fast Fourier transform of an orthogonal transform is applied. The output of the frequency component is performed in the form of a power spectrum. The output of the power spectrum is output numerically, and the distribution of the power spectrum may be judged by the degree of complexity by computer processing or the like, and colored so as to be easily recognized visually according to the strength of the power spectrum. May be represented by the following image.

The visual environment is evaluated by integrating the values of the VL and the power spectrum obtained above. The evaluation criterion may be represented by the visibility distance, or the visibility may be represented by three to five levels. Further, a method may be used in which one of the VL and the power spectrum is divided into several levels, and the other is used to determine the final visibility. Based on these evaluation results, a driver or the like may recognize a change in the visual environment by a sign or the like, and a gaze guide mark, a road stud, which is a self-luminous body installed around the road,
For example, the emission luminance may be increased in the case of heavy fog or a snowstorm to make it easier for a driver or the like to recognize.

The evaluation criteria are as follows under various environments.
The correlation between the measured contrast value and the value of the frequency distribution with respect to a visual environment obtained by actually recognizing a person is measured, and a threshold value is provided in a plurality of steps according to the degree of visibility.

According to the present invention, the visual environment can be evaluated in a manner much closer to human recognition than simply measuring the contrast, and the accuracy is high. Therefore, the present invention is used for recognition of dense fog and snowstorm on a road. For example, recognition of vehicles at intersections with poor visibility, recognition of traffic congestion by recognizing the presence of vehicles, security devices by recognition of intruders not only on roads, recognition of fire by sensing smoke, etc. It is also possible to apply.

[0022]

According to the present invention, since the surrounding optical conditions such as the brightness of the surroundings are measured by analyzing the frequency components in addition to the measurement of the contrast, a form close to human recognition can be obtained. Can be used to evaluate the visual environment. Since the measurement of the contrast and the measurement of the surrounding optical conditions are performed by using a digital image, the apparatus can be extremely simplified and easily made.

Further, since the measurement of the surrounding optical conditions is performed by analyzing the frequency components of the image, it is possible to analyze not only a part of the image but also a wide range. By using in combination, it is possible to highly evaluate the correlation with the visual environment to be visually recognized more accurately. In addition, since the evaluation can be performed over a wide range of the image, even if the target object changes due to contamination or the like, the evaluation can be performed as a whole, so that the influence of the change in the state of the target object on the evaluation can be reduced.

[Brief description of the drawings]

FIG. 1 is a diagram showing a flow of an example of a method for evaluating a visual environment according to the present invention.

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Ryuji Imazu 440-15 Okayama, Mikasa, Hokkaido Sekisui Jushi Co., Ltd. (72) Inventor Sei Kisaka 440-15 Okayama, Mikasa, Hokkaido Sekisui Jushi Co., Ltd. F-term (reference) 5B057 AA16 BA02 BA11 DA17 DB02 DB06 DB09 DC22 5H180 AA01 CC04 LL01 LL02 LL04 LL15 5L096 AA02 AA06 BA02 CA02 FA23 FA32

Claims (6)

[Claims] An image processing apparatus includes a digital image introducing unit and a digital image analyzing unit, and converts an image introduced by the digital image introducing unit into a visual environment by measuring a contrast obtained by the digital image analyzing unit and analyzing a frequency component. A visual environment evaluation method, which is used as a reference for judging a state. 2. The visual environment evaluation method according to claim 1, wherein the contrast measurement outputs a relative degree with respect to a previously obtained limit contrast. 3. The analysis of a frequency component according to claim 1, wherein the analysis of the frequency component is performed by converting a digital image into a frequency component by a fast Fourier transform of an orthogonal transform, and analyzing a distribution of a high frequency and a low frequency. Visual environment evaluation method. 4. The visual environment evaluation according to claim 1, wherein the analysis of the frequency component obtains a power spectrum for each digital image unit and obtains an average of the power spectra within a predetermined area. Method. 5. The image introduced by the digital image introduction means is divided into a plurality of areas, and the contrast measurement and the analysis of frequency components are performed by the digital image analysis means for each area. ,
The visual environment evaluation method according to 2, 3, or 4. 6. A self-luminous body, wherein light emission luminance is changed based on a visual environment state obtained by the visual environment evaluation method according to claim 1, 2, 3, 4, or 5.