KR102416820B1 - Method and Apparatus for detecting windows using building images - Google Patents

Abstract

The present invention relates to a window detection method and apparatus using a building image, the method comprising: setting an entire wall surface as a region of interest from an image of an exterior wall of a building having regularly arranged windows; calculating an average contrast value of each pixel based on each of a horizontal axis and a vertical axis of the region of interest; calculating a maximum value and a minimum value of the average intensity value with respect to the horizontal axis and the vertical axis, respectively, and setting a reference threshold value between the maximum value and the minimum value; and detecting a window region in the ROI using a set reference threshold value and the average intensity value.

Description

Method and Apparatus for detecting windows using building images}

The present invention relates to a method and apparatus for detecting a window using an image of a building, and more particularly, to a method and apparatus for detecting a window from an image of an exterior wall of a building having regularly arranged windows.

Recently, indoor monitoring technology for the elderly living alone has been researched to determine the movement of users and their stationed positions during indoor living, such as tracking the location using the distance and proximity sensors installed in the house or tracking the location using a switch sensor. is in progress

Existing indoor monitoring technology has the advantage of being able to know detailed emergency situations and preventing loneliness by analyzing life patterns.

However, there may be a problem of invasion of privacy due to the measurement of video and audio. In addition, the Internet is used to deliver the information obtained from the monitoring device or the situation information determined by processing the information. However, in an underdeveloped environment such as a rental apartment, the Internet connection itself may not be smooth, or communication costs may be reduced. Additional costs may be incurred for each household, which may put a burden on users economically.

Therefore, there is a need for an outdoor monitoring technology that can minimize the problem of privacy invasion, satisfy economic feasibility, and recognize the abnormal situation of the elderly or the disabled living alone in a limited environment.

The present invention is proposed in response to such a request, and in order to easily determine the situation of each household based on an image of the exterior wall of a building obtained outdoors with a photographing device such as CCTV, a technique for accurately determining each household is required.

Accordingly, one embodiment of the present invention is to provide a method and apparatus for accurately detecting each window from an image of an exterior wall of a building taken outdoors, and further determining each household based on the detected window.

The problem to be solved by the present invention is not limited to the problem(s) mentioned above, and another problem(s) not mentioned will be clearly understood by those skilled in the art from the following description.

A window detection method using an image of a building by an image processing apparatus according to an embodiment of the present invention includes: setting the entire wall as a region of interest from an image of an exterior wall of a building having regularly arranged windows; calculating an average contrast value of each pixel based on each of a horizontal axis and a vertical axis of the region of interest; calculating a maximum value and a minimum value of the average intensity value with respect to the horizontal axis and the vertical axis, respectively, and setting a reference threshold value between the maximum value and the minimum value; and detecting a window region in the ROI using a set reference threshold value and the average intensity value.

In one embodiment, the window detection method further includes determining a generation according to a preset criterion based on the detected window area.

In an embodiment, the step of setting the reference threshold value between the maximum value and the minimum value includes dividing the interval between the maximum value and the minimum value into preset sections for each of the horizontal axis and the vertical axis, and dividing the boundary value of each section as a candidate threshold value set to; classifying a plurality of single window regions and a single wall region in the region of interest based on the candidate thresholds for each of the horizontal and vertical axes; and selecting a reference threshold value from among a plurality of candidate thresholds based on lengths of the plurality of single window areas and single wall areas with respect to the horizontal axis and the vertical axis, respectively.

In an embodiment, the selecting of a reference threshold value from among the plurality of candidate threshold values includes defining a length of each of the plurality of single window areas as a first set, and defining a length of each of the plurality of single wall areas as a second set. defining a set; calculating a product of the variance of the first set and the variance of the second set for each of the candidate thresholds; and setting, as a reference threshold, a candidate threshold at which the product of the variance of the first set and the variance of the second set is minimized.

In an embodiment, between calculating the average intensity value of each pixel and setting a reference threshold value between the maximum value and the minimum value, the average intensity value of each pixel calculated using a linear regression equation is It characterized in that it further comprises the step of correcting the contrast difference.

In one embodiment, the preset criterion is characterized in that the number of windows corresponding to the first generation.

A window detection apparatus using a building image according to an embodiment of the present invention includes: a region of interest setting unit configured to set the entire wall as a region of interest from an image of an exterior wall of a building having regularly arranged windows; a contrast calculator configured to calculate an average intensity value of each pixel based on a horizontal axis and a vertical axis of the region of interest; a threshold value setting unit calculating a maximum value and a minimum value of the average intensity value with respect to the horizontal axis and the vertical axis, respectively, and setting a reference threshold value between the maximum value and the minimum value; and a window region detector configured to detect a window region in the ROI using a set reference threshold value and the average intensity value.

According to an embodiment of the present invention, it is possible to accurately detect a window in an image of an exterior wall of a building. In addition, by determining each household based on the detected window, it can be utilized for determining the abnormal state of each household in the future.

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

1 is a block diagram schematically illustrating the configuration of a device for detecting a window using a building image according to an embodiment of the present invention.
2 is a block diagram schematically illustrating a configuration of a controller of a device for detecting a window according to an embodiment of the present invention.
3 is a flowchart illustrating a window detection method using a building image according to an embodiment of the present invention.
4A is an example of an image of an exterior wall of a building according to an embodiment of the present invention, and FIG. 4B is an example of an ROI according to an embodiment of the present invention.
5 is a graph of an average intensity value of each pixel of an ROI according to an embodiment of the present invention.
6 is a flowchart illustrating a method for setting a reference threshold value according to an embodiment of the present invention.
7 is a diagram illustrating an example for explaining variance to which a candidate threshold is applied according to an embodiment of the present invention.
8 is an example of a graph to which a reference threshold is applied to a corrected average contrast graph according to an embodiment of the present invention.
9 is a two-dimensional image to which a reference threshold is applied to a region of interest according to an embodiment of the present invention.

Since the present invention can have various changes and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and it should be understood to include all modifications, equivalents and substitutes included in the spirit and scope of the present invention. In describing each figure, like reference numerals have been used for like elements.

Terms such as first, second, A, and B may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component. and/or includes a combination of a plurality of related listed items or any of a plurality of related listed items.

When an element is referred to as being “connected” or “connected” to another element, it is understood that it may be directly connected or connected to the other element, but other elements may exist in between. it should be On the other hand, when it is said that a certain element is "directly connected" or "directly connected" to another element, it should be understood that the other element does not exist in the middle.

The terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In the present application, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. does not

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically illustrating the configuration of a window detection device using a building image according to an embodiment of the present invention, and FIG. 2 is a configuration of a controller of the window detection device according to an embodiment of the present invention. It is a schematic block diagram.

Referring to FIG. 1 , an apparatus 100 for detecting a window using a building image is an image processing apparatus that acquires an image and processes the acquired image, and includes an input unit 110 , a display unit 120 , a communication unit 130 , and a storage unit. 140 and a controller 150 .

The input unit 110 converts a user's input operation into an input signal and transmits it to the controller 150 . The input unit 110 may be implemented as, for example, a keyboard, a mouse, a touch sensor on a touch screen, a touch pad, a keypad, a voice input, and other input processing devices capable of present, past, or future input processing. The input unit 110 may receive, for example, a user's information provision request input and transmit it to the controller 150 .

The display unit 120 outputs a screen according to the control of the controller 150 . The display unit 120 may be implemented as, for example, a liquid crystal display (LCD), a light emitting diode (LED), an organic light emitting diode (OLED), a projector, or other display devices capable of present, past or future. The display unit 120 may display, for example, a building image or an interface for setting an ROI in the building image.

The communication unit 130 exchanges data with the photographing device 10 and/or other external devices. The communication unit 130 transmits the image received from the photographing device 10 to the controller 150 . In addition, the communication unit 130 transmits data to other external devices, for example, a preset related organization according to the control of the controller 150 . The communication technology used by the communication unit 130 may vary depending on the type of communication network or other circumstances.

The storage unit 140 stores data under the control of the controller 150 and transmits the requested data to the controller 150 .

The controller 150 controls the overall operation of the window detection apparatus 100 and each component.

The controller 150 detects a window from an image of an exterior wall of a building, as will be described later. For example, the controller 150 displays an interface for setting a region of interest in the image of the exterior wall of a building obtained from the camera 10 through the display unit 120, detects a window in the region of interest set from the interface, and detects a result of the detection is displayed through the display unit 220 or transmitted to a preset external device and terminal through the communication unit 130 .

Referring to FIG. 2 , the controller 150 includes a region of interest setting unit 151 , a contrast calculation unit 152 , a correction unit 153 , a threshold value setting unit 154 , a window region detection unit 155 , and a generation determining unit. (156).

The region of interest setting unit 151 provides an interface for setting the region of interest in which the window region is to be detected from the image of the exterior wall of the building through the display unit 120, and the region of interest is selected from the user through the input unit 110 and is set. .

The contrast calculator 152 calculates a contrast value of each pixel in the set ROI, and obtains an average of the contrast value of each pixel with respect to each of the horizontal and vertical axes of the ROI. The shade of each layer can be known by the contrast of the horizontal axis.

The correction unit 153 corrects the average contrast value of each pixel by using a linear regression equation to reflect the illuminance value according to the influence of objects such as surrounding buildings or trees on the wall image.

The threshold value setting unit 154 sets a reference threshold value for classifying the window region and the wall region in the ROI based on the corrected maximum and minimum values of the average contrast value of each pixel. The reference threshold value setting will be described later with reference to FIGS. 4 to 8 .

The window region detection unit 155 detects a window region in the ROI using a set reference threshold value and the average intensity value.

The generation determining unit 156 determines the generation according to a preset criterion based on the detected window area. The preset criterion includes the number of windows corresponding to the first generation, and when the number of windows corresponding to the first generation is plural, further includes a window arrangement direction. For example, if the number of windows corresponding to one household is one building, the preset criterion includes one window in one household, and the household determination unit 156 determines each detected window as each household. do. When the number of windows corresponding to one household is two in the horizontal direction, the preset criterion includes two horizontal windows in one household, and the household determination unit 156 determines the neighbors of each floor (horizontal axis) from the detected windows. It is judged as one household by tying two windows that do.

These preset criteria may be input by an administrator.

The controller 150 of an embodiment of the present invention is implemented as a sum total of processing devices in which each part is physically separated, but some of each part of the controller 150 may be implemented in an integrated device, and the controller 150 All additions of may be implemented in one device.

A detailed operation of each component of the window detection apparatus 100 will be described later with reference to FIGS. 3 to 9 .

3 is a flowchart for explaining a window detection method using a building image according to an embodiment of the present invention, and FIG. 4 (a) is an example of an image of an exterior wall of a building according to an embodiment of the present invention, FIG. 4(b) is an example of a region of interest according to an embodiment of the present invention. 5 is a graph of an average intensity value of each pixel of an ROI according to an embodiment of the present invention.

Referring to FIGS. 3 and 4 , the entire wall is set as a region of interest from the image of the exterior wall of a building having regularly arranged windows ( S110 ).

The original image obtained from a camera or a photographing device such as a CCTV may include an exterior wall of a building having regularly arranged windows and other backgrounds, as shown in FIG. 4 ( a ). A user sets a region of interest from an image of an exterior wall of a building having regularly arranged windows through an interface provided by the controller. By selecting four vertices of the region set as the region of interest in the image of the exterior wall of the building, the region of interest as shown in FIG. 4(b) is set. The region of interest includes a plurality of regularly arranged windows in a grid pattern.

An average contrast value of each pixel is calculated based on each of the horizontal and vertical axes of the region of interest ( S120 ).

To do this, first, the contrast value of each pixel in the ROI is extracted. Next, the average of the brightness values of each pixel on the horizontal axis of the region of interest is calculated, and the average of the brightness values of each pixel on the vertical axis of the region of interest is calculated, and the average brightness value of each pixel is calculated based on each of the horizontal and vertical axes.

5A is a graph of the average intensity value of each pixel on the horizontal axis of the region of interest. The x-axis of the graph indicates the pixel on the horizontal axis of the region of interest, and the y-axis indicates the contrast value. 5B is a graph of the average intensity value of each pixel on the vertical axis of the region of interest. The x-axis represents the pixel on the vertical axis of the region of interest, and the y-axis represents the contrast value. Referring to (a) of Figure 5, it can be seen that the contrast (shade) of the horizontal axis corresponding to each floor of the building is periodically changed, and referring to Figure 5 (b), It can be seen that the contrast (shade) of each axis of the ROI on the vertical axis is periodically changed.

Referring back to FIG. 3 , the difference in brightness between the calculated average brightness values of each pixel is corrected using a linear regression equation ( S130 ).

In the original building exterior wall image, there may be differences in illuminance values for each area due to the influence of surrounding structures. Therefore, the effect of unwanted illuminance values is corrected in the extracted average contrast graph for each axis, and in this case, a cubic linear regression equation can be used. Since a correction method using a cubic linear regression equation is a conventional technique, a detailed description thereof will be omitted.

The maximum value and minimum value of the average brightness value of each pixel are calculated, and a reference threshold value is set between the maximum value and the minimum value (S140).

6 is a flowchart illustrating a method for setting a reference threshold value according to an embodiment of the present invention.

Referring to FIG. 6 , in order to set the reference threshold value, first, the interval between the maximum value and the minimum value of the average intensity value for each of the horizontal and vertical axes of the region of interest is divided into preset sections, and the boundary value of each section is set as a candidate threshold value. set (S142).

The preset section may be set by an administrator or a designer.

A plurality of single window areas and a single wall area are divided in the ROI based on the candidate thresholds for each of the horizontal and vertical axes (S144).

In the region of interest, each window area is referred to as a single window area, the single window area and the neighboring single window area are spaced apart from each other, and a wall area between the neighboring single window areas is referred to as a single wall area. That is, the single window area and the single wall area are alternately arranged.

Based on the lengths of the plurality of single window areas and single wall areas with respect to the horizontal axis and the vertical axis, respectively, a reference threshold value is selected from among a plurality of candidate threshold values as shown in Equation 1 (S146).

[Equation 1]

Here, Tmax is the maximum value of the average intensity value, and Tmin is the minimum value of the average intensity value. Di is the set of respective lengths of the single window area based on the i-th candidate threshold, and Ni is the set of respective lengths of the single wall area based on the i-th candidate threshold. V(Di) is the dispersion value of Di, and V(Ni) is the dispersion value of Ni.

For example, a length of each of the plurality of single window areas is defined as a first set (D), and a length of each of the plurality of single wall areas is defined as a second set (N). The lengths of the single window area and the single wall area can be confirmed in the average contrast graph for each axis. In the average contrast graph for each axis, each area separated based on the candidate threshold is a single window area and a single wall area. A region with a brightness value higher than the candidate threshold indicates a wall area, and a brightness value lower than the candidate threshold value is a wall area. The area represents the window area.

The product of the variance of the first set (D) and the variance of the second set (N) is calculated for each of the candidate thresholds. In other words, the product of the variance value V(Di) of the first set of lengths of the single window area of all candidate thresholds and the variance value V(Ni) of the second set of lengths of the single wall area is obtained.

A candidate threshold at which the product of the variance of the first set (D) and the variance of the second set (N) is minimized is set as the reference threshold. In other words, the smallest value among the products of V(Di) and V(Ni) of all the candidate thresholds sets a candidate threshold having a value of the product of V(Di) and V(Ni) as the reference threshold value.

7 is a diagram illustrating an example for explaining variance to which a candidate threshold is applied according to an embodiment of the present invention.

7A shows the first and second sets to which the first candidate threshold is applied, and FIG. 7B shows the first and second sets to which the second candidate threshold is applied.

Referring to FIG. 7 , the product of the variance of the first set of first candidate thresholds and the variance of the second set is 38.671, and the product of the variance of the first set and the second set of second candidate thresholds is 0.173. It can be seen that Fig. 7(b), in which the product of variance is small, has a more regular arrangement than in Fig. 7(a), in which the product of variance is large. That is, it can be confirmed that the smaller the product of variance is, the more regular the arrangement is.

For these results, the graph of FIG. 8 can be obtained by applying a threshold value after correcting the contrast graph of FIG. 5 .

8 is an example of a graph to which a reference threshold is applied to a corrected average contrast graph according to an embodiment of the present invention, and FIG. 9 is a two-dimensional image to which a reference threshold is applied to an ROI according to an embodiment of the present invention. to be.

FIG. 8(a) is a corrected average contrast graph on the horizontal axis, that is, the horizontal direction, of the region of interest, and FIG. 8(b) is a corrected average contrast graph on the vertical axis, that is, the vertical direction of the region of interest.

Referring again to FIGS. 3 and 8 , a window region is detected in the ROI using a set reference threshold value and the average intensity value ( S150 ).

For example, if the calculated reference threshold value is applied to the graph of FIG. 5 , the graph of FIG. 8 may be obtained.

If the graph of FIG. 8 is expanded to a two-dimensional image, the image of FIG. 9 (a) can be obtained, and when the image of FIG. 9 (a) is applied to the region of interest of the image of the exterior wall of the building of FIG. An image of the exterior wall of the building in which the window area is clearly distinguished as in (b) can be obtained.

A generation is determined according to a preset criterion based on the detected window area (S160).

Here, the preset criterion is the number of windows corresponding to the first generation. For example, when the number of windows corresponding to the first generation is 1, each window is determined as each household. When the number of windows corresponding to the first generation is 2, it is determined that the two windows are the 1st generation along the horizontal axis.

The relationship between windows and households may be different for each building. In other words, since the number of windows and the shape of windows in each building are different from each other, the standard for the relationship between households and windows should be set in advance by the manager.

Table 1 shows the evaluation results for the window detection method according to an embodiment of the present invention.

Truth Positive negative Detection Positive 326 93 Precision negative 16 0 77.80% Sensitivity Accuracy 95.32% 74.94%

Referring to Table 1, as a result of the experiment on 5 apartment wall images, as the detection results for a total of 342 windows, the value of classifying windows as windows is 326, the value of classifying windows as walls is 16, The value classified as a wall is 93, and the value classified as a wall is 0. Accordingly, it can be seen that the precision becomes 77.80%, the accuracy becomes 71.94%, the sensitivity becomes 95.32, and the F1-Score is 85.68%.

According to the window detection method according to an embodiment of the present invention, the window area is clearly detected from the image of the exterior wall of the building, and based on this, the household is determined or the lighting on/off state is accurately detected through the brightness of the window area, etc. It can be easily used for abnormal state determination and the like. In addition, according to the window detection method according to an embodiment of the present invention, since it is an outdoor measurement method, it helps to determine the abnormal state of the household without violating the user's privacy protection. In addition, according to the window detection method according to an embodiment of the present invention, it is possible to determine the abnormal state of the entire building rather than installing it for each household, so efficiency in management is expected.

Meanwhile, the embodiments of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the present invention, or may be known and available to those skilled in the art of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic such as floppy disks. - Examples of program instructions such as magneto-optical and ROM, RAM, and flash memory can be executed by a computer using an interpreter as well as machine code such as generated by a compiler. contains high-level language codes. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the embodiments of the present invention, and vice versa.

So far, the present invention has been looked at with respect to preferred embodiments thereof. Those of ordinary skill in the art to which the present invention pertains will understand that the present invention can be implemented in a modified form without departing from the essential characteristics of the present invention. Therefore, the disclosed embodiments are to be considered in an illustrative rather than a restrictive sense. The scope of the present invention is indicated in the claims rather than the foregoing description, and all differences within the scope equivalent thereto should be construed as being included in the present invention.

100: window detection device
110: input unit
120: display
130: communication department
140: storage
150: controller
151: region of interest setting unit
152: Contrast calculator
153: correction unit
154: threshold setting unit
155: window area detection unit
156: generation judging unit

Claims (12)

A window detection method using an image of a building by an image processing device, comprising:
setting the entire wall as a region of interest from an image of an exterior wall of a building having regularly arranged windows;
calculating an average contrast value of each pixel based on each of a horizontal axis and a vertical axis of the region of interest;
calculating a maximum value and a minimum value of the average intensity value with respect to the horizontal axis and the vertical axis, respectively, and setting a reference threshold value between the maximum value and the minimum value;
detecting a window region in the ROI using a set reference threshold value and the average intensity value; and
determining a household according to a preset criterion based on the detected window area;
including,
The preset criterion is the number of windows in the window arrangement direction corresponding to the first generation,
Setting a reference threshold value between the maximum value and the minimum value comprises:
dividing a section between the maximum value and the minimum value with respect to each of the horizontal and vertical axes into preset sections, and setting a boundary value of each section as a candidate threshold value;
classifying a plurality of single window regions and a single wall region in the region of interest based on the candidate thresholds for each of the horizontal and vertical axes; and
selecting a reference threshold value from among a plurality of candidate thresholds based on lengths of the plurality of single window areas and single wall areas with respect to the horizontal axis and the vertical axis, respectively;
The window detection method using a building image, characterized in that the single window represents each window area, and the single wall represents a wall area between the single window areas.
delete According to claim 1,
The window detection method using a building image, characterized in that the single window area and the single wall area are alternately arranged.
The method of claim 1,
The step of selecting a reference threshold value from among the plurality of candidate threshold values includes:
defining a length of each of the plurality of single window areas as a first set, and defining a length of each of the plurality of single wall areas as a second set;
calculating a product of the variance of the first set and the variance of the second set for each of the candidate thresholds; and
Setting a candidate threshold at which the product of the variance of the first set and the variance of the second set is the minimum as a reference threshold value
Window detection method using a building image, characterized in that it comprises a.
According to claim 1,
Between calculating the average intensity value of each pixel and setting a reference threshold value between the maximum value and the minimum value,
Compensating for the difference in contrast between the average contrast values of each pixel calculated using a linear regression equation
Window detection method using a building image further comprising a.
delete a region of interest setting unit for setting the entire wall as a region of interest from an image of an exterior wall of a building having regularly arranged windows;
a contrast calculator configured to calculate an average intensity value of each pixel based on a horizontal axis and a vertical axis of the region of interest;
a threshold value setting unit calculating a maximum value and a minimum value of the average intensity value with respect to the horizontal axis and the vertical axis, respectively, and setting a reference threshold value between the maximum value and the minimum value;
a window region detector configured to detect a window region in the region of interest using a set reference threshold value and the average contrast value; and
A household determination unit that determines a household according to a preset criterion based on the detected window area
including,
The preset criterion is the number of windows in the window arrangement direction corresponding to the first generation,
The threshold setting unit,
For each of the horizontal and vertical axes, a section between the maximum value and the minimum value is divided into preset sections, a boundary value of each section is set as a candidate threshold, and the region of interest is based on the candidate threshold value for each of the horizontal and vertical axes. distinguishing a plurality of single window areas and single wall areas, and selecting a reference threshold value from among a plurality of candidate thresholds based on lengths of the plurality of single window areas and single wall areas with respect to the horizontal and vertical axes, respectively;
The window detection apparatus using a building image, characterized in that the single window represents each window area, and the single wall represents a wall area between the single window areas.
delete delete 8. The method of claim 7,
The threshold setting unit,
A length of each of the plurality of single window areas is defined as a first set, and a length of each of the plurality of single wall areas is defined as a second set, and the variance of the first set and the second set for each of the candidate thresholds are calculating the product of the variance of the set, and setting a candidate threshold at which the product of the variance of the first set and the variance of the second set is the minimum as a reference threshold
Window detection device using a building image, characterized in that.
8. The method of claim 7,
A correction unit that corrects the difference between the brightness and the average brightness value of each pixel calculated using a linear regression equation
Window detection device using a building image further comprising a.





delete

Images