KR20220126474A - Apparatus and method for detecting defect pixel based on multi-band images - Google Patents

Abstract

According to a preferred embodiment of the present invention, provided are a device and a method for detecting a defective pixel based on multi-band images. The method may detect a defective pixel candidate group in the spatial domain by reflecting the characteristics of each multi-band, and detect the final defective pixel in consideration of the amount of change in the time domain based on the defective pixel candidate group in the spatial domain, so as to be applicable to systems using multi-band images, and additionally detect the defective pixel in consideration of image characteristics of each band. Therefore, it is not necessary to separately store the defective pixel map for each band in the memory.

Description

Apparatus and method for detecting defect pixel based on multi-band images}

The present invention relates to an apparatus and method for detecting a bad pixel based on a multi-band image, and more particularly, to an apparatus and method for detecting a bad pixel in a multi-band image.

The infrared image searcher has an infrared sensor that measures the amount of radiation from an object with an absolute temperature of 0°C or higher. The precision of the measured data is proportional to the sensitivity of the sensor, and the higher the sensitivity, the more accurately and precisely the irradiance of the measuring object can be measured. Since the temperature resolution and accuracy of the infrared image searcher used as the missile's warhead is one of the most important hardware specifications, much effort is needed to improve the sensor's performance.

Looking at the recent development trend of the infrared image searcher, as the performance of the sensor is stabilized, the development of searchers that apply multiple sensors (multi-band) to compensate for the shortcomings of the system using it rather than improving hardware performance is a trend.

However, in the case of the conventional multi-sensor (multi-band), two methods of calculating bad pixels by utilizing the characteristics of a single sensor (single-band) are used. One method is to check the output deviation of each pixel using an image obtained while pointing to a surface having uniform radiation characteristics, and to calculate pixels having a difference greater than or equal to a specified value as bad pixels. Another method is to observe a change amount of an image continuously input and observe the values of pixels determined not to be a target to calculate a bad pixel.

The purpose of using a multi-band image is to select and borrow only the advantages of each band using a fusion image. For example, the pixel value at position (10,10) does not satisfy the criteria for bad pixels in the characteristics of the image acquired in each band and is used as a normal pixel. can be calculated. That is, when only the calculation criteria in the single-band image are applied to the fusion image, there is a problem in that it is impossible to calculate additionally possible bad pixels.

Therefore, it is necessary to develop a method for detecting a bad pixel by using a multi-band image.

An object of the present invention is to detect a bad pixel candidate group in a spatial domain by reflecting multi-band characteristics, and to detect a final bad pixel in consideration of a change amount in the time domain based on the bad pixel candidate group in the spatial domain. , to provide an apparatus and method for detecting a bad pixel based on a multi-band image.

Other objects not specified in the present invention may be additionally considered within the scope that can be easily inferred from the following detailed description and effects thereof.

According to a preferred embodiment of the present invention for achieving the above object, there is provided an apparatus for detecting a bad pixel based on a multi-band image, comprising: an image acquisition unit configured to acquire a multi-band image; A sum image and a difference image are obtained based on the multi-band image for a current frame obtained through the image acquisition unit, and a preset edge detection filter is used based on the sum image and the difference image. a first detector configured to acquire a noisy image by using a first threshold value preset based on the noise image to detect a group of bad pixel candidates of the current frame; and a second detector configured to detect a final bad pixel of the current frame by using a change amount of a candidate bad pixel in a time domain based on the group of bad pixel candidates of the current frame detected by the first detector.

Here, the second detection unit may acquire the candidate bad pixel in the group of the bad pixel candidates in which an amount of change in the time domain of the candidate bad pixel is greater than a preset second threshold as the final bad pixel.

Here, the second detection unit obtains a first adjustment value by applying a preset first weight to an output value of the candidate bad pixel for a frame immediately preceding the current frame, for each of the candidate bad pixels in the bad pixel candidate group; , by applying a second weight obtained by subtracting the first weight from 1 to the output value of the candidate bad pixel for the previous frame of the immediately preceding frame to obtain a second adjustment value, the first adjustment value and the second adjustment value obtains a predicted value of the candidate bad pixel with respect to the current frame based on It may be determined whether the candidate bad pixel is the final bad pixel based on the 2 threshold value.

Here, the first detection unit applies the edge detection filter to the sum image to obtain the filtered sum image, applies the edge detection filter to the difference image to obtain the filtered difference image, The noise image may be obtained by subtracting the filtered difference image from the sum image.

Here, the edge detection filter may be a Roberts mask-based filter.

Here, the Roberts mask may include a first mask consisting of {(-1, 0), (0, 1)} and a second mask consisting of {(0, -1), (1, 0)} have.

According to a preferred embodiment of the present invention for achieving the above object, there is provided a method for detecting a bad pixel based on a multi-band image, the method comprising: acquiring a multi-band image; Obtaining a sum image and a difference image based on the multi-band image for a current frame, and obtaining a noise image using a preset edge detection filter based on the sum image and the difference image, detecting a bad pixel candidate group of the current frame by using a first threshold preset based on the noise image; and detecting a final bad pixel of the current frame by using a change amount of a candidate bad pixel in a time domain based on the bad pixel candidate group of the current frame.

Here, the step of detecting the final bad pixel may include acquiring, as the final bad pixel, the candidate bad pixel whose change amount in the time domain of the candidate bad pixel is greater than a preset second threshold in the bad pixel candidate group. .

Here, the final bad pixel detection step includes applying a first weight preset for each of the candidate bad pixels in the bad pixel candidate group to the output value of the candidate bad pixel for the frame immediately preceding the current frame to obtain a first adjustment value. and applying a second weight obtained by subtracting the first weight from 1 to the output value of the candidate bad pixel with respect to the previous frame of the immediately preceding frame to obtain a second adjustment value, the first adjustment value and the second weight obtaining a predicted value of the candidate bad pixel with respect to the current frame based on an adjustment value, and obtaining a difference value between the predicted value of the candidate bad pixel and an output value of the candidate bad pixel with respect to the immediately preceding frame, and the difference value and It may be determined whether the candidate bad pixel is the final bad pixel based on the second threshold value.

Here, the step of detecting the bad pixel candidate group includes applying the edge detection filter to the sum image to obtain the filtered sum image, and applying the edge detection filter to the difference image to obtain the filtered difference image, The noise image may be obtained by subtracting the filtered difference image from the filtered sum image.

Here, the edge detection filter may be a Roberts mask-based filter.

Here, the Roberts mask may include a first mask consisting of {(-1, 0), (0, 1)} and a second mask consisting of {(0, -1), (1, 0)} have.

A computer program according to a preferred embodiment of the present invention for achieving the above technical object is stored in a computer-readable recording medium, and any one of the above-described multi-band image-based bad pixel detection methods is executed in the computer.

According to an apparatus and method for detecting a bad pixel based on a multi-band image according to a preferred embodiment of the present invention, a bad pixel candidate group in the spatial domain is detected by reflecting the characteristics of each multi-band, and time based on the bad pixel candidate group in the spatial domain By detecting the final bad pixel in consideration of the amount of change in the region, it can be applied to a system using a multi-band image, and bad pixels can be additionally detected in consideration of the image characteristics for each band, and a bad pixel map for each band is separately stored in memory. You don't have to save it to .

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a block diagram illustrating an apparatus for detecting a bad pixel based on a multi-band image according to a preferred embodiment of the present invention.
2 is a diagram for explaining a process of detecting a bad pixel in a spatial domain according to a preferred embodiment of the present invention.
3 is a diagram for explaining a process of detecting a bad pixel in a time domain according to a preferred embodiment of the present invention.
4 is a flowchart illustrating a method for detecting a bad pixel based on a multi-band image according to a preferred embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments published below, but may be implemented in various different forms, and only these embodiments make the publication of the present invention complete, and common knowledge in the art to which the present invention pertains It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout.

Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used with the meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless clearly defined in particular.

In the present specification, terms such as “first” and “second” are for distinguishing one component from other components, and the scope of rights should not be limited by these terms. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

In the present specification, identification symbols (eg, a, b, c, etc.) in each step are used for convenience of description, and identification symbols do not describe the order of each step, and each step is clearly Unless a specific order is specified, the order may differ from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

In this specification, expressions such as “have”, “may have”, “include” or “may include” indicate the presence of a corresponding feature (eg, a numerical value, function, operation, or component such as a part). and does not exclude the presence of additional features.

In addition, the term '~ unit' as used herein means software or a hardware component such as a field-programmable gate array (FPGA) or ASIC, and '~ unit' performs certain roles. However, '-part' is not limited to software or hardware. '~unit' may be configured to reside on an addressable storage medium or may be configured to refresh one or more processors. Thus, as an example, '~' denotes components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data structures and variables. The functions provided in the components and '~ units' may be combined into a smaller number of components and '~ units' or further separated into additional components and '~ units'.

Hereinafter, a preferred embodiment of an apparatus and method for detecting a bad pixel based on a multi-band image according to the present invention will be described in detail with reference to the accompanying drawings.

First, an apparatus for detecting a bad pixel based on a multi-band image according to a preferred embodiment of the present invention will be described with reference to FIGS. 1 to 3 .

1 is a block diagram illustrating a multi-band image-based bad pixel detection apparatus according to a preferred embodiment of the present invention, and FIG. 2 is a block diagram for explaining a process of detecting a bad pixel in a spatial domain according to a preferred embodiment of the present invention. 3 is a diagram for explaining a process of detecting a bad pixel in a time domain according to a preferred embodiment of the present invention.

Referring to FIG. 1 , a multi-band image-based bad pixel detection apparatus (hereinafter, referred to as a 'bad pixel detection apparatus') 100 according to a preferred embodiment of the present invention reflects multi-band characteristics for bad pixels in a spatial domain. A candidate group is detected, and a final bad pixel is detected in consideration of a change amount in the temporal domain based on the bad pixel candidate group in the spatial domain.

To this end, the bad pixel detection apparatus 100 may include an image acquisition unit 110 , a first detection unit 130 , and a second detection unit 150 .

The image acquisition unit 110 may acquire a multi-band image.

That is, the image acquisition unit 110 may include a plurality of sensors (not shown) that measure the amount of radiation in different bands, and may acquire a multi-band image through the plurality of sensors. For example, the output of a sensor measuring radiation energy in different bands has different output values even if the same target is indicated.

The first detector 130 may detect a bad pixel candidate group in the spatial domain by reflecting the multi-band characteristics.

That is, referring to FIG. 2 , the first detection unit 130 performs a multi-band image (the first band in FIG. 2 ) for the current frame (the N-th frame in FIG. 2 ) acquired through the image acquisition unit 110 . A sum image and a difference image may be obtained based on the image frame to the nth band image frame).

In addition, the first detector 130 may acquire a noise image using a preset edge detection filter based on the sum image and the difference image.

In more detail, the first detector 130 applies the edge detection filter to the sum image to obtain a filtered sum image, applies the edge detection filter to the difference image to obtain a filtered difference image, and the filtered sum image A noise image may be obtained by subtracting the filtered difference image from .

Here, the edge detection vector may be a Roberts mask-based filter. The Roberts mask may include a first mask made of {(-1, 0), (0, 1)} and a second mask made of {(0, -1), (1, 0)}.

In addition, the first detector 130 may detect a bad pixel candidate group of the current frame (the Nth frame of FIG. 2 ) using a first threshold preset based on the noise image.

In other words, an image of WxH size obtained through sensor A capable of measuring radiant energy for band A is called “A_image”, and an image of size WxH obtained through sensor B capable of measuring radiant energy for band B is used. Assuming "B_image", the sum and difference of images acquired at the same time in each band can be obtained to examine the post-fusion characteristics of two images. The reason for calculating the sum and difference of the acquired images is to determine the noise generated by the primary target.

Bad pixels of images acquired using two different sensors exist in different positions. However, by obtaining the sum of the two images, the output deviation of the target and the background is increased by increasing the image output value of the target part, and the difference in the output values across the images, that is, the noise, can be obtained by calculating the difference between the two images.

Sum_image = A_image + B_image

Here, Sum_image represents a sum image.

Sub_image = A_image - B_image

Here, Sub_image represents a difference image.

By applying a 2x2 size Roberts mask-based filter to each of the sum image and the difference image, a filtered sum image and a filtered difference image in which band-specific characteristics are reflected may be generated. By comparing the filtered sum image and the filtered difference image, the positions of the candidate bad pixels can be found by excluding the positions of the edge components caused by the target. Thereafter, pixels that do not pass the first threshold value among the corresponding candidate bad pixels may be determined as the bad pixel candidate group detected in the spatial domain. Here, the 2x2 Roberts mask has a relatively high speed, and it can be used because an edge that is clearly distinguished from the surrounding is detected and noise is more prominent than other filters.

mask1 = {(-1, 0), (0, 1)}

Here, mask1 represents the first mask of the Roberts mask.

mask2 = {(0, -1), (1, 0)}

Here, mask2 represents the second mask of the Roberts mask.

Sum_image_filtered = (Sum_image * mask1) + (Sum_image * mask2)

Here, Sum_image_filtered represents a filtered sum image generated by applying a Roberts mask to the sum image.

Sub_image_filtered = (Sub_image * mask1) + (Sub_image * mask2)

Here, Sub_image_filtered represents a filtered difference image generated by applying a Roberts mask to the difference image.

Image_noise = Sum_image_filtered - Sub_image_filtered

Here, Image_noise represents a noise image.

The second detector 150 may detect the final bad pixel in consideration of the amount of change in the time domain based on the candidate group of bad pixels in the spatial domain detected by the first detector 130 .

That is, referring to FIG. 3 , the second detection unit 150 detects a bad pixel candidate group (candidate bad pixels 1 to 3 in FIG. 3 ) of the current frame (the Nth frame of FIG. 3 ) detected by the first detection unit 130 . Based on pixel n), the final bad pixel (last bad pixel 1 to last bad pixel m in FIG. 3) of the current frame (Nth frame in FIG. 3) can be detected by using the amount of change in the time domain of the candidate bad pixel have.

In more detail, the second detector 150 may acquire, as a final bad pixel, a candidate bad pixel whose change amount in the time domain of the candidate bad pixel is greater than a preset second threshold from the bad pixel candidate group.

In this case, the second detection unit 150 may determine whether the corresponding candidate bad pixel is the final bad pixel for each of the candidate bad pixels in the bad pixel candidate group.

That is, the second detector 150 applies a preset first weight to the output value of the candidate bad pixel with respect to the frame immediately preceding the current frame (the Nth frame of FIG. 3 ) (the N−1th frame of FIG. 3 ) to obtain the first Adjustment values can be obtained.

Then, the second detector 150 applies the second weight to the output value of the candidate bad pixel with respect to the previous frame (the N-2 th frame of FIG. 3 ) of the immediately preceding frame (the N-1 th frame of FIG. 3 ) to obtain a second Adjustment values can be obtained. Here, the second weight represents a value obtained by subtracting the first weight from 1 .

In addition, the second detector 150 may obtain a predicted value for the current frame (the N-th frame of FIG. 3 ) of the candidate bad pixel based on the first adjustment value and the second adjustment value.

Then, the second detector 150 detects a difference between the predicted value of the candidate bad pixel for the current frame (the Nth frame of FIG. 3 ) and the output value of the candidate bad pixel with respect to the previous frame (the N−1th frame of FIG. 3 ). can be obtained

In addition, the second detector 150 may determine whether the candidate bad pixel is the final bad pixel based on the difference value and the second threshold value.

In other words, it is difficult to clearly determine a bad pixel due to the target and the background only by the process detected in the spatial domain. Accordingly, it is possible to additionally determine whether a pixel is a bad pixel by using the images acquired in the time domain.

In an additional determination process, the present invention may apply a weight value k for each time component. That is, the value k is applied to the frame (N-1 th frame) immediately before the input current frame (N-th frame), and (1) to the previous frame (N-2 th frame) of the immediately preceding frame (N-1 th frame). By applying the -k) value, a value in consideration of the amount of change in the existing output value of the pixel can be calculated. Thereafter, the deviation I from the previous frame (the N−1th frame) is calculated to calculate the amount of change in the output value of the pixel, and when the I value exceeds a specific second threshold value, the pixel may be determined as a final bad pixel.

Nth_img = k * (N-1)th_img + (1-k) * (N-2)th_img

Here, Nth_img represents a predicted value of the Nth frame of the candidate bad pixel. k represents the first weight. (N-1)th_img represents an output value of a candidate bad pixel for an N-1 th frame. (1-k) represents the second weight. (N-2)th_img represents an output value of a candidate bad pixel for an N-2 th frame.

I = Nth_img - (N-1)th_img

Here, I denotes a difference value between candidate bad pixels.

Next, a method for detecting a bad pixel based on a multi-band image according to a preferred embodiment of the present invention will be described with reference to FIG. 4 .

4 is a flowchart illustrating a method for detecting a bad pixel based on a multi-band image according to a preferred embodiment of the present invention.

Referring to FIG. 4 , the bad pixel detection apparatus 100 may acquire a multi-band image ( S110 ).

Then, the bad pixel detection apparatus 100 acquires a sum image and a difference image based on the multi-band image of the current frame, and acquires a noise image using a preset edge detection filter based on the sum image and the difference image, , a bad pixel candidate group of the current frame may be detected using a preset first threshold based on the noisy image ( S130 ).

In this case, the bad pixel detection apparatus 100 applies the edge detection filter to the sum image to obtain a filtered sum image, applies the edge detection filter to the difference image to obtain a filtered difference image, and filters the filtered sum image. A noise image may be obtained by subtracting the difference image. Here, the edge detection vector may be a Roberts mask-based filter. The Roberts mask may include a first mask made of {(-1, 0), (0, 1)} and a second mask made of {(0, -1), (1, 0)}.

Then, the bad pixel detection apparatus 100 may detect the final bad pixel of the current frame by using the change amount of the candidate bad pixel in the time domain based on the bad pixel candidate group of the current frame ( S150 ).

That is, the bad pixel detection apparatus 100 may obtain, as a final bad pixel, a candidate bad pixel whose change amount in the time domain of the candidate bad pixel is greater than a preset second threshold in the bad pixel candidate group.

In this case, the bad pixel detection apparatus 100 applies a preset first weight to the output value of the candidate bad pixel for the frame immediately preceding the current frame to obtain a first adjustment value for each of the candidate bad pixels in the bad pixel candidate group, A second adjustment value is obtained by applying the second weight (the value obtained by subtracting the first weight from 1) to the output value of the candidate bad pixel for the previous frame of the previous frame, and based on the first adjustment value and the second adjustment value Obtaining the prediction value of the current frame of the candidate bad pixel, obtaining the difference value between the predicted value of the current frame of the candidate bad pixel and the output value of the candidate bad pixel with respect to the previous frame, and obtaining a candidate based on the difference value and the second threshold It may be determined whether the bad pixel is a final bad pixel.

Even though all the components constituting the embodiment of the present invention described above are described as being combined or operated in combination, the present invention is not necessarily limited to this embodiment. That is, within the scope of the object of the present invention, all of the components may operate by selectively combining one or more. In addition, all of the components may be implemented as one independent hardware, but some or all of the components are selectively combined to perform some or all functions of the combined components in one or a plurality of hardware program modules It may be implemented as a computer program having In addition, such a computer program is stored in a computer readable media such as a USB memory, a CD disk, a flash memory, etc., read and executed by a computer, thereby implementing an embodiment of the present invention. The computer program recording medium may include a magnetic recording medium, an optical recording medium, and the like.

The above description is merely illustrative of the technical idea of the present invention, and various modifications, changes and substitutions are possible within the scope that does not depart from the essential characteristics of the present invention by those of ordinary skill in the art to which the present invention pertains. will be. Accordingly, the embodiments disclosed in the present invention and the accompanying drawings are for explaining, not limiting, the technical spirit of the present invention, and the scope of the technical spirit of the present invention is not limited by these embodiments and the accompanying drawings. . The protection scope of the present invention should be construed by the following claims, and all technical ideas within the equivalent range should be construed as being included in the scope of the present invention.

100: bad pixel detection device;
110: image acquisition unit,
130: a first detection unit;
150: second detection unit

Claims (13)

an image acquisition unit for acquiring a multi-band image;
A sum image and a difference image are obtained based on the multi-band image for a current frame obtained through the image acquisition unit, and a preset edge detection filter is used based on the sum image and the difference image. a first detector configured to acquire a noisy image by using a first threshold value preset based on the noise image to detect a group of bad pixel candidates of the current frame; and
a second detector configured to detect a final bad pixel of the current frame by using a change amount of a candidate bad pixel in a time domain based on the group of bad pixel candidates of the current frame detected by the first detector;
A multi-band image-based bad pixel detection device comprising a. In claim 1,
The second detection unit,
obtaining, as the final bad pixel, the candidate bad pixel having a change amount of the candidate bad pixel in the time domain greater than a preset second threshold in the bad pixel candidate group;
Multi-band image-based bad pixel detection device. In claim 2,
The second detection unit,
For each candidate bad pixel in the bad pixel candidate group, a first adjustment value is obtained by applying a preset first weight to the output value of the candidate bad pixel for the frame immediately preceding the current frame, and the first weight is A second adjustment value is obtained by applying the subtracted second weight to the output value of the candidate bad pixel for the previous frame of the immediately preceding frame, and based on the first adjustment value and the second adjustment value, obtain a prediction value for the current frame, obtain a difference value between the predicted value of the candidate bad pixel and an output value of the candidate bad pixel with respect to the immediately preceding frame, and obtain the candidate based on the difference value and the second threshold value determining whether the bad pixel is the final bad pixel;
Multi-band image-based bad pixel detection device. In claim 1,
The first detection unit,
The edge detection filter is applied to the sum image to obtain the filtered sum image, the filtered difference image is obtained by applying the edge detection filter to the difference image, and the filtered difference image is obtained from the filtered sum image. To obtain the noise image by subtracting the image,
Multi-band image-based bad pixel detection device. In claim 4,
The edge detection filter is
Roberts mask based filter,
Multi-band image-based bad pixel detection device. In claim 5,
The Roberts mask,
A first mask consisting of {(-1, 0), (0, 1)} and a second mask consisting of {(0, -1), (1, 0)},
Multi-band image-based bad pixel detection device. acquiring a multi-band image;
Obtaining a sum image and a difference image based on the multi-band image for a current frame, and obtaining a noise image using a preset edge detection filter based on the sum image and the difference image, detecting a bad pixel candidate group of the current frame by using a first threshold preset based on the noise image; and
detecting a final bad pixel of the current frame by using a change amount of a candidate bad pixel in a time domain based on the bad pixel candidate group of the current frame;
A multi-band image-based bad pixel detection method comprising a. In claim 7,
The final bad pixel detection step includes:
and acquiring, as the final bad pixel, the candidate bad pixel whose change amount in the time domain of the candidate bad pixel is greater than a preset second threshold in the bad pixel candidate group;
Multi-band image-based bad pixel detection method. In claim 8,
The final bad pixel detection step includes:
For each candidate bad pixel in the bad pixel candidate group, a first adjustment value is obtained by applying a preset first weight to the output value of the candidate bad pixel for the frame immediately preceding the current frame, and the first weight is A second adjustment value is obtained by applying the subtracted second weight to the output value of the candidate bad pixel for the previous frame of the immediately preceding frame, and based on the first adjustment value and the second adjustment value, obtain a prediction value for the current frame, obtain a difference value between the predicted value of the candidate bad pixel and an output value of the candidate bad pixel with respect to the immediately preceding frame, and obtain the candidate based on the difference value and the second threshold value It consists of determining whether the bad pixel is the last bad pixel,
Multi-band image-based bad pixel detection method. In claim 7,
The step of detecting the bad pixel candidate group includes:
The edge detection filter is applied to the sum image to obtain the filtered sum image, the filtered difference image is obtained by applying the edge detection filter to the difference image, and the filtered difference image is obtained from the filtered sum image. It consists of obtaining the noise image by subtracting the image,
Multi-band image-based bad pixel detection method. In claim 10,
The edge detection filter is
Roberts mask based filter,
Multi-band image-based bad pixel detection method. In claim 11,
The Roberts mask,
A first mask consisting of {(-1, 0), (0, 1)} and a second mask consisting of {(0, -1), (1, 0)},
Multi-band image-based bad pixel detection method. A computer program stored in a computer-readable recording medium in order to execute the method for detecting a bad pixel based on a multi-band image according to any one of claims 7 to 12 in a computer.

Images