KR100290099B1 - Image inputting device having a wide dynamic range and thereof method - Google Patents

Abstract

PURPOSE: A device for inputting images having a dynamic range of a wide band and a method thereof are provided to obtain maximum image information without any data loss by processing analog input signals with a plurality of A/D converters, and obtain clear image of a subject by making histogram uniform adaptively according to an average luminance of divided areas of the histogram. CONSTITUTION: A device for inputting images having a dynamic range of a wide band includes a charge-coupled device(1) for taking an image of a subject to output a corresponding analog image, a signal conversion part(4) having a plurality of level shifters elements for shifting signal level of the analog image signal by a set level and a plurality of A/D converters for converting analog image signals output from the level shifters to digital image signals, thereby outputting a plurality of digital image signals corresponding to the analog image signal output from the charge-coupled device, a digital data producing part(5) for extracting valid digital data from the plurality of digital image signals, and a histogram control part(6) for extracting a histogram from the valid digital data, obtaining areas on an image coordinates system correspondingly to divided histogram areas if the extracted histogram has an area able to be divided by a luminance value, and carrying out histogram processing for the obtained areas.

Description

Image input device having wide dynamic range and method thereof {IMAGE INPUTTING DEVICE HAVING A WIDE DYNAMIC RANGE AND THEREOF METHOD}

The present invention relates to an image input apparatus having a wide dynamic range and a method thereof, and more particularly, to an apparatus for converting, processing, storing, and transmitting an analog input signal into a digital signal. The present invention provides a video input device and a method thereof having a wide dynamic range for processing a plurality of analog / digital converters to obtain maximum image information.

A device that converts an analog input signal into a digital signal, processes, stores, and transmits. A device that converts an audio signal into a digital signal includes a digital recorder and an audio device. The device converts an image signal into a digital signal. Surveillance digital CCD cameras, digital still cameras, and digital movie cameras.

In general, the dynamic range of data capable of representing analog signals corresponding to captured images as digital data having a predetermined length is extremely limited. Accordingly, the following methods have been used so far to reproduce images with better quality.

First, it transforms the shape of the analog input signal itself, and in the case of an image signal, the shape of the signal using the principle (Weber ratio) that the difference between large values is much more noticeable than the difference between small values. This is a similar way to log conversion.

However, in this method, there is a problem in that the overall width of the brightness level in the image is very wide and the effect is not obtained when it is concentrated in a certain area.

The second method is to process the histogram of digital image data that has already been acquired to increase the width of the signal as a whole, and to express the characteristics of the image signal in terms of the frequency of magnitude (in the case of image data, the histogram). It is called histogram equalization in the sense of uniformly distributing the histogram by a method of widening the difference relatively to the concentrated part.

However, the second method has a problem in that its performance is limited by using a value converted into digital data having a limited number already.

In addition, by ignoring the area information of the subjects actually present in the image, when there is a dark area and a bright area, and the difference is large, the detailed image inside each area is not distinguished, so the effect is remarkably inferior. That is, when the brightness difference between the background and the subject is remarkable, such as backlit photographing by the camera, since the histogram smoothing is performed throughout the picture, a problem that the subject or the background cannot be made more prominent occurs.

As mentioned above, conventional methods for widening the signal width by converting the shape of an analog signal or manipulating the value of a signal that has already been digitally converted also cause the following problems.

That is, the method of converting the form of the analog signal has a problem that it is impossible to represent all the various signal forms by distorting the signal form of the input terminal.

In addition, a method of manipulating a value that has already been converted to a digital signal has a problem in that a lot of information is actually lost in the process of converting to a limited number of digital values, and the lost value cannot be reproduced through any subsequent processing.

Accordingly, the technical problem of the present invention is to solve such a conventional problem, and an object of the present invention is to segment an area of a histogram extracted from image data obtained from an analog input signal, and adaptively according to the average brightness of each divided area. By uniformizing the histogram, an image input device having a wide dynamic range for sharpening a subject image for each region is provided.

Another object of the present invention is to provide an image input method having a wide dynamic range using the image input apparatus having the obtained wide dynamic range.

1 is a block diagram illustrating an image input apparatus having a wide dynamic range in an embodiment of the present invention.

2 is a block diagram of an A / D conversion unit according to an embodiment of the present invention;

3 is an operation flowchart of an image data histogram method of an image input apparatus having a wide dynamic range according to an embodiment of the present invention;

4 is a state diagram showing data distribution according to A / D conversion according to an embodiment of the present invention,

5A and 5B are histogram distribution diagrams according to an embodiment of the present invention.

An image input apparatus having a wide dynamic range according to one feature for realizing the above object of the present invention,

An imaging unit for capturing an image of a subject and outputting a corresponding analog image signal;

A plurality of level shifters for moving the signal level of the analog video signal output from the image pickup unit according to a set level, and a plurality of A / Ds for converting each analog video signal output from the plurality of level shifters into a digital video signal; A signal converter having a converter and outputting a plurality of digital video signals by converting a plurality of analog video signals output from the imaging unit;

A digital data generator for extracting valid digital data from a plurality of digital video signals output from the signal converter; And

Extracting a histogram from valid digital data output from the digital data generator, and in the case where a segmentable region exists according to a brightness value on the extracted histogram, an area in an image coordinate system corresponding to each region of the divided histogram is obtained. And a histogram adjusting unit which performs histogram processing for each of the obtained regions.

In addition, the video input method having a wide dynamic range according to one feature for realizing the above invention,

(a) converting an analog image signal corresponding to the photographed subject into a plurality of digital image signals;

(b) extracting valid digital data from the converted plurality of digital video signals;

(c) extracting a histogram from the valid digital data;

(d) checking whether or not a segmentable area exists according to the brightness value on the extracted histogram;

(e) obtaining a region on an image coordinate system corresponding to each region of the divided histogram when there is a segmentable region in step (d); And

and (f) performing a histogram extension process for each of the obtained regions.

According to the image input device and the method having such a wide dynamic range, the analog input signal is processed by a plurality of analog / digital converters to obtain the maximum image information without data loss, and the histogram according to the obtained plurality of digital data. By extracting and dividing the region, and adaptively uniformizing the histogram according to the average brightness of each divided region, the subject image can be made clear for each region.

With the above configuration, the most preferred embodiment which can be easily carried out by those skilled in the art to the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram illustrating an image input apparatus having a wide dynamic range in an embodiment of the present invention.

2 is a block diagram of an A / D conversion unit according to an embodiment of the present invention;

3 is an operation flowchart of an image data histogram method of an image input apparatus having a wide dynamic range according to an embodiment of the present invention;

4 is a state diagram showing data distribution according to A / D conversion according to an embodiment of the present invention,

5A and 5B are histogram distribution diagrams according to an embodiment of the present invention.

As shown in FIG. 1, a configuration of an image input apparatus having a wide dynamic range according to an embodiment of the present invention is provided.

An optical system 1 for forming an image of a subject, an image capturing unit 2 mounted on an output side of the optical system 1, and outputting an analog signal corresponding to an image formed on an object; and an output terminal of the image capturing unit 2 A signal processor 3 for removing noise of an analog signal and adjusting a signal gain, and multiple A / Ds for converting and outputting an analog signal applied to an output terminal of the signal processor 3 into a corresponding digital signal. A high precision digital data generation unit 5 connected to an output unit of the multiple A / D conversion unit 4 and outputting valid data from an applied digital signal; and a high precision digital data generation unit. A histogram that extracts the histogram from the effective digital data applied to the output terminal of (5) and divides the region, and adaptively uniforms the histogram according to the average brightness of each divided region. It consists of the adjustment part 6.

As shown in FIG. 2, the multiple A / D converter 4 includes a plurality of level shifters for moving a level of an applied signal connected to an output terminal of the signal processor 3 according to different reference levels. L1 to Ln and a plurality of A / D converters 41 to 4n, n = 1, 2, 3, ... which are connected to output terminals of the plurality of level shifters L1 to Ln, respectively.

The operation of the image input apparatus having the wide dynamic range according to the embodiment of the present invention by the above configuration is as follows.

In the present invention, to improve the A / D process of acquiring digital data, a plurality of existing A / D converters are used to obtain a lot of information from a limited number of digital data.

However, the following stages of image processing related to the processing, storage and transmission of digital image data generally consist of the precision of the output of a single A / D converter or less. Accordingly, a process of reducing information of the data is performed while taking maximum information from a large amount of data obtained.

The process of reducing the information of the data is based on the conventional image histogram manipulation method, but unlike the conventional method, by using the very high degree of digital data obtained from the outputs of a plurality of digital converters, the brightness distribution is manipulated as it is. As a result, image data having maximum information can be obtained.

First, when power is applied to an image processing apparatus such as a digital still camera, system initialization is performed and an imaging unit 2 is imaged through the optical system 1 corresponding to a subject to be photographed.

The image pickup unit 2 outputs an electrical signal corresponding to an image-shaped subject, that is, an analog signal, and the analog signal is input to the signal processing unit 3.

The signal processing section 3 performs the same signal processing operation as before, to remove noise of the input analog signal, and to adjust the signal gain appropriately.

As described above, the analog signal from which the noise is removed by the signal processing unit 3 is input to the multiple A / D converter 4 and converted into a corresponding digital signal and output.

The analog signal input to the multiple A / D converter 4 is output as a plurality of analog signals having different reference levels, respectively, through a plurality of level shifters L1 to Ln, as shown in FIG. and a plurality of digital data are output through the n A / D converters 41 to 4n.

For example, when a digital signal between 0 to 4V is to be digitally converted through an A / D converter having 0 to 2V input and 10 bit output, a level shifter having a reference level of 0V and a reference of 2V A level shifter having a level is used, and a signal output through each level shifter is digitally converted through a 10-bit A / D converter and output. The result is a total of 20 bits of digital data with improved precision.

In the present invention, the number of n, i.e., the number of level shifters and A / D converters, is adjusted according to the required data precision.

In the above, the plurality of A / D converters 41 to 4n process analog signals having different levels, thereby providing overlapping portions to prevent data loss between two adjacent converters. There are three cases of output data of two adjacent converters as described in Table 1 below.

The plurality of digital data output from the multiple A / D converter 4 is input to the high precision digital data generator 5, and the high precision data generator 5 is valid digital data from the applied digital data. Extract the data and print it out.

Table 1 below shows an example in which an arbitrary analog signal is digitally converted.

Converter input n n + 1 Transducer n + 1 or higher level 0x10 0xFF Nesting level 0x04 0xF3 Transducer n or less level 0x00 0xD6

As shown in FIG. 4, in the case of digitally converting the A signal corresponding to the converter n + 1 level or higher, the n A / D converter outputs a valid value of '0x10' as shown in Table 1 above. The n + 1 A / D converter outputs an invalid value of '0xFF'. When the B signal corresponding to the overlap level is digitally converted, a valid value of '0x04' is output by the n A / D converter, and a valid value of '0xF4' is output by the n + 1 A / D converter. Is output. In addition, when the C signal corresponding to the converter n level or lower is digitally converted, the n A / D converter outputs an invalid value of '0x00', but the n + 1 A / D converter outputs '0xD6'. Outputs a valid value of.

Accordingly, the high-precision digital data is a compensation value from the '0x04' and '0xF4' digital data corresponding to the overlap level, in order to calculate the actual data for the valid data '0x10' and '0xD6' from the plurality of values. δm is calculated. In other words,

The data (n) is compensated digital data, the data is digital data to be compensated for, and the compensation value δm is calculated as follows.

That is, the compensation value δm is an average value of compensation values between a plurality of overlapping data. For example, when the digital data corresponding to the overlapping level are '0x04' and '0xF3', respectively, the compensation value δ is calculated as follows. do.

δ = 4 + (0xFF-0xF3)

Valid digital data processed by the high precision digital generator 5 as described above is input to the histogram adjusting unit 6 and processed.

The histogram adjusting unit 6 according to the present invention uses an adaptive histogram transformation technique that takes into account the spatial characteristics of the histogram smoothing.

That is, the biggest problem in the conventional histogram smoothing is that the spatial relationship between the pixels is not taken into consideration. In the present invention, the histogram is adaptively uniformized according to the average brightness of each region in consideration of the spatial characteristics between the pixels.

As shown in FIG. 3, the histogram adjusting unit 6 first extracts a histogram from the input digital image data (S110).

In general, a histogram of digital image data having a brightness level of [0, L-1] may be expressed as a discrete function as follows.

R _k denotes the k-th brightness level, n _k denotes the number of pixels having the brightness level k, and n denotes the total number of pixels included in the image data. That is, the histogram p (r _k ) means the probability that the brightness level r _k will appear inside the image.

The input histogram is extracted by processing the input digital image data according to the discrete function described above, and then the extracted histogram is divided (S120).

More specifically, when the histogram is extracted according to the histogram extraction, the histogram region is divided according to the threshold values TH1 and TH2 of the brightness values when there are histograms with significantly different brightness values.

When there is a histogram region with a significantly different brightness value as described above, the region on the image coordinate system corresponding to the set histogram region is found.

The area segmentation method of the present invention is performed by limiting a range on a histogram and applying the same to an image coordinate system.

A range in which brightness values exist on a histogram of a given image is defined, and n histogram regions overlapping each other are previously set. In operation S140, an area on the image coordinate system included in the set histogram area is found.

In the above, an area on the histogram represents a range of image brightness values, and an area on an image coordinate system represents an area of one two-dimensional closed curve that can be divided into brightness values in the image.

For example, when the camera is back-lit, the subject comes out black and the background is taken brightly, so that the shape of the subject is emphasized. However, in the related art, by performing smoothing on the histogram extracted over the entire screen, the difference between the brightness values of the subject and the background is reduced, thereby reducing the effect of backlighting.

According to the present invention, the histogram extracted over the entire screen is divided into regions according to the difference in brightness values, the region corresponding to the coordinate system of the image corresponding to the divided arbitrary histogram region is obtained, and the histogram is performed for each obtained region. It is.

For example, as shown in FIG. 5A, when the set histogram area is equal to or smaller than the threshold TH2 and equal to or greater than TH1, the peripheral pixel of the pixel a included between the threshold TH1 and TH2 in the arbitrary pixel a is the set value. (Eg, 1/2 of the total number of peripheral pixels), the pixel a is determined to be included in the set histogram area.

The above operation is applied to all pixels to find an area on the image coordinate system corresponding to the histogram area that is equal to or smaller than the threshold TH2 and equal to or greater than TH1. The region on the image coordinate system corresponding to the histogram region that is equal to or greater than the threshold TH2 or the region that is equal to or less than the threshold TH1 may be found by the same method.

Therefore, the arbitrary image is divided into one two-dimensional closed curved area that corresponds to the set histogram area and can be divided into brightness values in the image.

For example, at the time of back light imaging of the camera, the image is divided into two-dimensional closed curved areas corresponding to a subject having a low brightness value and two-dimensional closed curved areas corresponding to a background.

Next, histogram processing is performed on each area of the divided image as described above (S150).

Therefore, the histogram for each region inside the image is individually adjusted so that the maximum value of the histogram is equal to the maximum value of the brightness value, and the minimum value coincides with the minimum value of the brightness value, thereby expanding the histogram as shown in FIG. 5B. Is processed.

According to the above operation, the histogram is uniformly uniformed according to the average brightness of each region in consideration of spatial characteristics.

However, if there is no region that can be distinguished by the brightness value on the histogram extracted as shown in FIG. 5A, the histogram smoothing of the entire region is performed in the same manner as in the prior art (S160).

As described above, in the image processing apparatus for converting an analog signal into a digital signal and processing the same according to an embodiment of the present invention, the analog input signal is processed by a plurality of analog / digital converters, so that the maximum image information can be obtained without data loss. Can be obtained.

In addition, by dividing the histogram according to the digital data, which is the maximum image information, and dividing the region, and adaptively uniformizing the histogram according to the average brightness for each divided region, it is possible to make the image of the subject distinct for each region.

In addition, the histogram is adaptively uniformized according to the brightness of each region of the subject and the background when the camera photographs the backlight, thereby emphasizing the shape of the subject more clearly.

Claims (8)

An imaging unit for capturing an image of a subject and outputting a corresponding analog image signal; A plurality of level shifters for moving the signal level of the analog video signal output from the image pickup unit according to a set level, and a plurality of A / Ds for converting each analog video signal output from the plurality of level shifters into a digital video signal; A signal converter having a converter and outputting a plurality of digital video signals by converting a plurality of analog video signals output from the imaging unit; A digital data generator for extracting valid digital data from a plurality of digital video signals output from the signal converter; And Extracting a histogram from valid digital data output from the digital data generator, and in the case where a segmentable region exists according to a brightness value on the extracted histogram, an area in an image coordinate system corresponding to each region of the divided histogram is obtained. Histogram adjustment unit that performs histogram processing for each Image input device having a wide dynamic range comprising a. The method of claim 1, When the signal converter is composed of n (n = 1, 2, ...) level shifters and A / D changers, A signal output from an n + 1 A / D converter adjacent to any n A / D converter has a wide dynamic range characterized by being able to output one of a converter n + 1 level, an overlap level, and a converter n level or below. Video input device. The digital data generator of claim 1, wherein the digital data generator comprises: To obtain valid data from the digital video signal output from the signal converter And a compensation process using a formula of data (n) = data + 255 (n-1)-δm). The method of claim 1, wherein the histogram adjustment unit, If there is no segmentable area according to the brightness value on the extracted histogram, the image input device having a wide dynamic range, characterized in that to smooth the extracted histogram to improve the image contrast. The method of claim 1, wherein the histogram adjustment unit, In the arbitrary pixel on the image coordinate system, when the peripheral pixel of the arbitrary pixel included in the already divided | segmented setting range corresponding to the divided histogram area is more than a setting value, the said arbitrary pixel is divided into the area | region on the image coordinate system corresponding to the divided | segmented arbitrary histogram area | region. An image input device having a wide dynamic page, which is determined to be included. (a) converting an analog image signal corresponding to the photographed subject into a plurality of digital image signals; (b) extracting valid digital data from the converted plurality of digital video signals; (c) extracting a histogram from the valid digital data; (d) checking whether or not a segmentable area exists according to the brightness value on the extracted histogram; (e) obtaining a region on an image coordinate system corresponding to each region of the divided histogram when there is a segmentable region in step (d); And (f) performing a histogram extension process for each obtained region Image input method having a wide dynamic range comprising a. The method of claim 6, If there is no segmentable region according to the brightness value on the histogram extracted in the step (d), smoothing the extracted histogram to improve the image contrast; Image input method with dynamic range. 7. The method of claim 6, wherein extracting valid digital data comprises: To obtain valid data from digital video signal Compensation using the formula data n = data + 255 (n-1) -δm, where data n is the compensated digital video signal, data is the digital video signal to be compensated, and δm is the compensation value. A video input method having a wide dynamic range characterized by processing.

Images