JPH0236322A - Device and method for processing image - Google Patents

Abstract

PURPOSE:To accurately measure the intensity of incident light all over the measured range by selecting (1/2)N (N is an integer>=0) for the damping ratio of a light damping filter and multiplying a signal extracted from an output from an image input means by 2N at the time of obtaining intensity information corresponding to the above-mentioned ratio. CONSTITUTION:Light from an observed object is made incident on the image input means 2 through the light damping filter 1 having (1/2)N damping ratio. The output from the means 2 is inputted in a processing part 3 constituted of a microcomputer etc., and sampled every picture element, then converted into signal values 0-63 in 64 stages, for example, which are displayed as the digital value of 6 bits. The means 2 can output the signal at a level corresponding to the level of the incident light in the ratio of 1:1 only in the case of making the light having intensity in a specified range incident and otherwise it outputs the signal of a certain value. Therefore, the signal value of the picture element obtained by making the light having the intensity equal to or below the lower limit of the specified range incident becomes '0' and the signal value of the picture element obtained by making the light having the intensity equal to or above the upper limit thereof incident becomes '63'.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention is applicable to cases in which optical quantity distributions such as illuminance and brightness are measured using an image input means such as a CCD (charge-coupled device) camera. Image processing that is suitable for
v! The present invention relates to an image processing method and an image processing method.

[Conventional technology]

Traditionally, for example, when evaluating or considering lighting facilities, a so-called COD (charge-coupled device) camera or the like is used to capture an image of the observation target, and the output of this CCD camera is subjected to digital image processing to measure the pattern of brightness and darkness. An image processing apparatus is used that obtains distributions of optical quantities such as illuminance and brightness. In such devices, the output signal of the camera (CCI) is 64 steps, 256 steps, or 10 steps.
It is evaluated as a signal value of 24 levels or the like.

[Problem to be solved by the invention]

FIG. 3 is a graph showing changes in the output signal with respect to the intensity of incident light in the CCD camera. Generally, in a CCD camera, etc., a signal with a level that corresponds one-to-one (does not necessarily correspond linearly) to the incident light intensity level is produced only in a certain specific range as shown by the reference mark W1 in FIG. cannot be output. That is, reference mark W2. W3
For light intensities in the range indicated by , the signal is at a constant level, and therefore the light intensity cannot be measured.

For example, when measuring the light intensity distribution shown in FIG. 4(1), the output signal of the CCD camera becomes as shown in FIG. 4(2). That is, although it is ideal for the output signal of the CCD camera to change as shown by the curve 11 in FIG. 4(2), the output signal overflows in the range shown by the reference mark Pi, It reaches a certain level as shown by the solid line. Therefore, the above-mentioned signal value obtained from this CCD camera output also does not change within the range indicated by the reference mark P1, which means that the light intensity distribution in this range cannot be measured.

In order to solve this problem and expand the range, we added a so-called auto-range function that uses a neutral density filter (ND filter) and aperture to attenuate the intensity of the incident light in response to the intensity of the incident light. Devices are conventionally used. Such devices generally have a damping ratio of 101
1 (N is an integer greater than or equal to zero) and 57, and the value of N is changed based on the magnitude of the signal value.

For example, the CCD camera output can be set to 64 levels of signal values 0 to 6.
If the incident light intensity is distributed in the range corresponding to signal values 0 to 630 in the device evaluated in step 3,
By setting N to 1 and the damping ratio to 1/10, C
It is possible to input light with an intensity corresponding to a signal value in the range of 0 to 63 into a CD camera, and the resulting CCD
By multiplying the signal value based on the output signal of the camera by the reciprocal of the attenuation ratio lO, the light intensity distribution in the range corresponding to the signal value 0 to 630 can be measured without causing the above-mentioned overflow.

However, when the damping ratio is set to 1/1O, CC
The signal value l based on the D camera output corresponds to the signal value 10 of 15 cases where the attenuation ratio is 1/1, so the attenuation ratio is set to 1/1.
In this case, the change in the light intensity distribution in the range corresponding to the signal values 1 to 10 cannot be measured, and a problem arises in that the number of effective digits of the intensity information in this range decreases.

Furthermore, is it 31 measurements of the intensity of the attenuated light? ), the calculation for determining the intensity information of the incident light before attenuation is performed by a processing device equipped with, for example, a microcomputer, and since the calculation at this time is a binary calculation, the attenuation ratio (
If it is set to 1/10)', there is a problem that such arithmetic processing becomes complicated.

An object of the present invention is to provide an image processing bag and an image processing method that can accurately measure the intensity of incident light distributed over a wide range over the entire range and simplify the calculation of intensity information. be.

[Means to solve the problem]

The image processing device of the present invention includes a plurality of neutral density filters having mutually different attenuation ratios that attenuate the intensity of incident light with an attenuation ratio of (1/2) N, where N is an integer greater than or equal to zero; an image input means for guiding light from the filter and outputting a signal having a level corresponding to the intensity level in a ratio of 1 to 1 in response to incident light having an intensity in a certain specific range; means for extracting the output of the image input means corresponding to each neutral density filter, the level not exceeding the level corresponding to the upper limit of the specific range;
means for extracting a signal between an upper limit level of the extracted signal associated with a neutral density filter having an attenuation ratio N smaller than that neutral density filter; and each of said attenuation ratios (1/2) N. means for calculating intensity information of incident light by multiplying the signal extracted for each neutral density filter by 2N; and means for calculating intensity distribution of incident light by combining the intensity information obtained for each neutral density filter. Equipped with.

Further, in the image processing method of the present invention, N is an integer greater than or equal to zero, and N is an integer greater than or equal to zero, and the image input means outputs a signal having a level corresponding to an intensity level of 1 to 1 in response to incidence of light having an intensity in a certain specific range. When the intensity of the incident light is attenuated by an attenuation ratio of (1/2)y', a plurality of neutral density filters having mutually different attenuation ratios are successively interposed to guide the incident light, and the input light is ．． By means of extracting the output of the image input means corresponding to the light for each neutral density filter, the image input means output corresponding to the light is extracted for each neutral density filter. Extract the signal between the upper limit level of the signal extracted in relation to the optical filter, and multiply the signal extracted for each neutral density filter with the attenuation ratio (1/2)'' by 2N to obtain the incident light. The intensity distribution of the incident light is calculated by combining the intensity information obtained for each neutral density filter.

[For production]

According to the configuration of this invention, the image input means includes (l/2
)N (where N is an integer greater than or equal to zero), incident light is guided through a plurality of neutral density filters having mutually different attenuation ratios.

The image input means can output a signal having a level corresponding to the intensity level 1 to 1 only when light having an intensity in a certain specific range is incident. In the output of such an image input means, a signal that does not exceed an output signal level corresponding to the upper limit of the specific range is extracted every time a likelihood filter is inserted. However, signals extracted in connection with a neutral density filter having an attenuation ratio N smaller than that neutral density filter are not extracted. By multiplying this extracted signal by 2N', which is the reciprocal of the attenuation ratio of each neutral density filter,
It is possible to obtain a strong 'ttff information corresponding to the light intensity of the incident light before passing through the neutral density filter. This strength information is
As a result of the image input means output extraction processing as described above, for low intensity incident light, the image input means output corresponding to the neutral density filter with a small value of N is used, and for high intensity incident light, the value of N is Since it is determined based on the output of the image input means corresponding to the neutral density filter having a large value, the intensity information can have the same number of effective digits for the incident light intensity in the entire measurement range.

By combining the intensity information for each of the neutral density filters obtained in this way, it is possible to obtain the intensity distribution of the incident light over the entire region to be measured.

In this invention, the attenuation ratio of the neutral density filter is selected to be (1/2)H, and correspondingly, when obtaining intensity information, the signal extracted from the output of the image input means is multiplied by 2N. There is. For example, in a device that performs binary operations such as a microcomputer, such operations are performed by processing data in N
Bit shifting is sufficient, so the calculation of intensity information does not become unnecessarily complicated.

〔Example〕

FIG. 1 is a conceptual diagram showing the basic configuration of an embodiment of the present invention. This device measures the distribution of optical quantities such as brightness of an observation target illuminated by the lighting facility in order to evaluate and study the lighting facility. The light from the observation target is passed through a neutral density filter (ND filter) which has an attenuation ratio of (l/2)N (where N is an integer less than zero), using a CCD (charge-coupled device) camera as an example. The image enters the image input means 2. The output from the image input means 2 is input to a processing unit 3 that includes a microcomputer, memory, etc., and is sampled for each pixel to obtain a signal value of 64 levels expressed as a 6-bit digital value, for example. ~63.

Since the image input means 2 uses an imaging element such as a CCD, it is only possible to output a signal with a level that corresponds one-to-one to the incident light of a certain specific range of intensity. , the output signal remains at a constant level in a range other than this. Therefore, the signal value of a pixel into which light with an intensity below the lower limit of the specific range is incident becomes 0, and the signal value of a pixel into which light with an intensity above the upper limit is incident is 63. In this embodiment, a plurality of neutral density filters 1 corresponding to N=0.2.4.6 are sequentially replaced and used.

FIG. 2 is a conceptual diagram showing the signal processing procedure in this embodiment. The original image to be observed has, for example, a light intensity distribution in the range of 0 to 3840 in terms of signal values,
Assume that the distribution of light intensity is measured as the signal value of 0 to 3840.

The observation by the image input means 2 for such an original image is 2° (-1/1) and 22 (= 1/4) 2', respectively.
This is performed for each of the neutral density filters 11, 12, 14, and 16 (referred to as "neutral density filter l" when referred to as &8) having an attenuation ratio of (-1/16) and 2'' (-1/64). As a result of the incident light intensity being attenuated by each neutral density filter l, the received light intensity of the image input means 2 is converted into a signal value of O to 3840.0 for each neutral density filter 11.1214.16.
The range is from 9600 to 240.0 to 60.

Since the image input means 2 can only respond to the received light intensity within the above-mentioned specific range, the signal values read through the image processing board 4 range from 0 to 63 for the neutral density filters 11° and 12.14. Regarding the filter 16, it is 0 to 60. In other words, all pixels on which light with an intensity corresponding to a signal value of 63 or more is incident overflow and all have a signal value of 6.
It is considered to be 3.

Such signal values are stored in a memory provided within the processing device 3. Memory Lii in this memory is the second
In the figure, storage areas shown in a simplified manner corresponding to the original image, for example, storage areas indicated by reference symbols al and a2a4 are storage areas where the above-mentioned overflow has occurred and, therefore, the signal value 63 is stored.

First, the signal value obtained by observation with the neutral density filter 11 interposed is extracted in the processing section 3 together with the address information in the memory mentioned above, with the range of 0 to 59 being regarded as valid data, for example.

The extraction area corresponding to the original image at this time is the diagonally shaded area indicated by reference numeral b1 in FIG. The reason why the signal value near the signal value 63 is not extracted is due to the reliability in the vicinity of the upper limit of the specific range of the image input means.

Next, the signal values obtained using the neutral density filter 12 are extracted, with values from 15 to 59, for example, being regarded as valid data. Note that the range of signal values below 14 is extracted in relation to the dark filter 11. The extraction area corresponding to the original image at this time is the diagonally shaded area indicated by reference numeral b2.

Similarly, signal values in the range 15 to 59 are extracted for the neutral density filter 14, and signal values in the range 15 to 60 are extracted for the neutral density filter 16. These extraction regions are respectively referenced b4. b6
This is the shaded area indicated by .

The signal value extracted for each neutral density filter 1 in this way is multiplied by the reciprocal of the attenuation ratio of each neutral density filter 1, and then the signal value is determined from 0 to 3840 based on the address information in the memory of each signal value. are synthesized as an image matrix represented by . That is, neutral density filters 11, 12, 14, 16
2° (−
1), 22(= 41.2'(= 16), 2'(-6
4) is multiplied. Such operations can be easily performed by shifting the data by 0.2, 4.6 bits, respectively, within the processing section 3. The signal value after this calculation corresponds to the intensity of the incident light before passing through the neutral density filter l, so it can be called intensity information, and this intensity information is synthesized by the processing unit 3. The aforementioned image matrix obtained will correspond to the light intensity distribution in the original image.

In the intensity information corresponding to each of the neutral density filters 11, 12, 14.16, the pitch between the intensity information is 1.4.16.64 when expressed as a signal value. In addition, the range of intensity information corresponding to each neutral density filter 11, 12, 14.16 is 0 to 59.60 to 236, 24 when expressed in signal values.
0~944°960~3840. Therefore, according to this embodiment, the ratio of the pitch between intensity information to the intensity information is uniform over the entire range of signal values O to 3840, and therefore the effective number of digits of the intensity information is changed over the entire measurement range. It becomes possible to make it almost uniform. This makes it possible to measure the intensity distribution with a wide dynamic range and with high precision.

In the embodiment described above, four neutral density filters 11, 1214 .
16 and calculate each of these damping ratios as (l/2)N(1/2)
2. (l/2)N and (1/2)6, but the number of neutral density filters is arbitrary, and the attenuation ratio can be set to (1/2)'' to be different for each neutral density filter. However, other values may be used.

〔Effect of the invention〕

According to the image processing device and image processing method of the present invention,
Incident light attenuated by an attenuation ratio of (l/2)N is guided to an image input means, and the output of this image input means does not exceed an output signal level corresponding to the upper limit of a specific range that changes in response to the received light intensity. A signal is extracted, and intensity information is calculated by multiplying the extracted signal by 2N. However, in the extraction process, signals extracted with respect to an attenuation ratio N smaller than the attenuation ratio at which the extraction is performed are not extracted. Such processing is performed for a plurality of different N's, and then the intensity information obtained for each N is combined to calculate the intensity distribution of the incident light.

The interval between the values of the intensity information is defined by the value of N,
Furthermore, as a result of the above-described extraction process, intensity information is determined based on the output of the image input means, which corresponds to a small N for low intensity and a large N for high intensity, so that the interval between the values of the intensity information is determined based on the output of the image input means. The ratio of the value of the intensity information to the value of the intensity information is approximately uniform over the entire range of the obtained intensity information.

This makes it possible to obtain intensity information with the same number of significant digits over the entire measurement range, making it possible to accurately measure the intensity of incident light over the entire measurement range. become.

Further, in this invention, the attenuation ratio of the neutral density filter is (l/2
)N, and correspondingly, when obtaining intensity information, 2N is selected for the signal extracted from the output of the image input means.
I try to multiply it by

For such an operation, in a device such as a microcomputer that performs binary operations, it is sufficient to shift the data by N bits, so that the operation of the intensity information does not become unnecessarily complicated. This greatly simplifies the calculation of the intensity information.

[Brief explanation of the drawing]

Fig. 1 is a conceptual diagram showing the basic configuration of an embodiment of the present invention, Fig. 2 is a conceptual diagram showing its signal processing procedure, and Fig. 3 shows changes in the output signal with respect to the intensity of incident light in a CCD camera. The graph in FIG. 4 is a diagram for explaining the characteristics of a CCD camera. 1.11, 12.14.16...Dark filter, 2.
...Image input means, 3...Processing unit □IP Takesho

Claims (2)

[Claims] (1) A plurality of neutral density filters with mutually different attenuation ratios that attenuate the intensity of incident light with an attenuation ratio of (l/2)^N, where N is an integer greater than or equal to zero, and light from the neutral density filters. an image input means for outputting a signal with a level corresponding one-to-one to the intensity level in response to incident light having an intensity in a certain specific range; and an image corresponding to the incident light passing through each neutral density filter. means for extracting the output of the input means for each neutral density filter, the level not exceeding the level corresponding to the upper limit of the specific range;
means for extracting a signal between the level of the upper limit of the extracted signal associated with a neutral density filter having an attenuation ratio N smaller than that of the neutral density filter; A means for calculating the intensity information of the incident light by multiplying the signal extracted for each dark filter by 2^N, and a means for calculating the intensity information of the incident light by combining the intensity information obtained for each dark filter. An image processing device comprising: means for calculating. (2) When N is an integer greater than or equal to zero (l/2) for an image input means that outputs a signal with a level that corresponds one-to-one to the intensity level when light with an intensity in a certain range is incident. ^
A plurality of neutral density filters having mutually different attenuation ratios that attenuate the intensity of the incident light with an attenuation ratio of A level that does not exceed a level corresponding to the upper limit of the specific range by means of extracting for each neutral density filter;
extracting a signal between the upper limit level of the extracted signal associated with a neutral density filter having an attenuation ratio N smaller than that neutral density filter; The signal extracted for each optical filter is multiplied by 2^N to calculate the intensity information of the incident light, and the intensity information obtained for each neutral density filter is combined to calculate the intensity distribution of the incident light. Featured image processing method.