KR20190102736A - Method and Apparatus for Video Denoising Considering Moving Object - Google Patents

Abstract

Disclosed are a method of image noise reduction and an apparatus therefor in consideration of a moving object.
Determining camera movement between frames of an image, estimating background motion and moving body motion based on camera movement, and comparing the flow profile of each background motion and moving body motion with reference frame in the area where moving body exists The present invention relates to a method for removing video noise considering a moving object and a device therefor.

Description

Image Noise Reduction Method and Apparatus Therefor {Method and Apparatus for Video Denoising Considering Moving Object}

Embodiments of the present invention relate to an image noise removing method and apparatus therefor considering a moving object.

The contents described in this section merely provide background information on the embodiments of the present invention and do not constitute a prior art.

In general, image noise reduction techniques may be classified into a method of removing image noise in a spatial domain and a method of removing image noise in a time domain.

As a technique for removing noise in an image in the spatial domain, a method of performing filtering on the spatial domain is mainly used. The filtering method in the spatial domain includes a method of removing noise in consideration of a pattern of image noise (Korean Patent Laid-Open No. 10-2015-014259), and a method of removing noise in consideration of sensor noise characteristics (Korean Patent Publication No. 10-2007-0061309).

In addition, a technique for removing image noise in the time domain is a method of removing image noise by setting a reference image and adjusting the filtering intensity according to the degree of change of the screen determined using the reference image and the surrounding image. 10-2004-0024888) are mainly used.

However, since the above-described methods remove image noise using only the change of the background pixel without considering the motion of the camera or the moving object, distortion is generated even if the filtering intensity is adjusted. Therefore, there is a need for a method for accurately removing image noise in consideration of camera movement or moving object movement.

An embodiment of the present invention determines a camera motion between frames of an image, estimates a background motion and a moving motion based on the movement of the camera, and references a flow profile for each of the background motion and the moving motion in a region where the moving object exists. A main object of the present invention is to provide an image noise removing method and an apparatus therefor in consideration of a moving object for removing image noise.

According to an aspect of an embodiment of the present invention, in the apparatus for removing image noise in consideration of a moving object, the scene change information of the input image is extracted, and the divided image is divided by dividing the image by the same scene based on the scene change information. An input image processor to generate; A motion estimator for checking a movement of a camera in the divided image and estimating a background motion and a moving body motion based on the movement of the camera; And an image noise remover configured to remove image noise in an area in which a moving object exists by using a flow profile of the background motion and the moving object motion.

According to another aspect of an embodiment of the present invention, in the method for removing image noise in consideration of a moving object, the scene change information of the input image is extracted, and the image is divided and divided by the same scene based on the scene change information. An input image processing process of generating an image; A motion estimation process of checking a movement of a camera in the divided image and estimating a background motion and a moving body motion based on the movement of the camera; And an image noise removing process of removing image noise in an area in which a moving object exists by using a flow profile of the background motion and the moving object motion.

As described above, according to the exemplary embodiment of the present invention, it is possible to obtain a high quality image noise elimination result regardless of the movement of the camera sensor and the image noise pattern.

In addition, according to an embodiment of the present invention, by eliminating the video noise by using partial image information on the moving object region, it is possible to shorten the time for removing the video noise and to apply to the streaming environment.

In addition, according to an embodiment of the present invention, the video noise reduction technology can be applied to the streaming environment, there is an effect that can be applied to various fields such as IPTV, real-time streaming content, real-time monitoring of intelligent video system.

1 is a block diagram schematically illustrating an apparatus for removing video noise according to an exemplary embodiment of the present invention.
2 is a flowchart illustrating a method for removing image noise according to an exemplary embodiment of the present invention.
3 is an exemplary diagram for describing an operation of removing image noise according to an exemplary embodiment of the present invention.
4 is an exemplary view for explaining an operation of estimating a motion of a background and a moving object in an image noise removing device according to an exemplary embodiment of the present invention.
5 is an exemplary diagram for describing an operation of matching an image and removing image noise in an image noise removing apparatus according to an exemplary embodiment of the present invention.

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

1 is a block diagram schematically illustrating an apparatus for removing video noise according to an exemplary embodiment of the present invention.

The image noise canceller 100 according to the present exemplary embodiment includes an input image processor 110, a motion estimator 120, and an image noise remover 130.

The input image processor 110 obtains an input image, determines whether the input image is switched, and divides the input image into segments for the same space. The input image may be a video image captured by a camera and may be an image captured by a plurality of spatial regions. For example, the input image processor 110 may acquire an image of photographing a plurality of preset regions using an Pan-Tilt-Zoom (PTZ) camera as an input image.

The input image processor 110 extracts scene change information of the input image and divides the scene change information into images captured in a continuous space using the scene change information. The input image processor 110 describes the divided image based on the same space as the divided image, and in this embodiment, the noise is removed by using a frame included in the divided image.

The input image may be implemented as one divided image when there is no spatial change, that is, a scene change, and may be implemented as a plurality of divided images when there is a spatial change.

The input image processor 110 determines a scene change of the input image by using the scene change information, and specifically, a scene is changed to a frame showing a sudden change based on the scene change information including color histogram, feature point correspondence accuracy, and the like. By selecting a frame, it is possible to determine whether to change scenes.

The input image processor 110 may determine whether the scene changes based on scene change information included in metadata information of the input image, but is not limited thereto. Difference) It is also possible to determine whether to change scenes by selecting a frame whose error is greater than or equal to a threshold. For example, assuming that the input image is an image composed of 100 frames photographing the A region and the B region, the input image processing unit 110 extracts the 50th frame from which the photographing of the B region is started based on the scene change information. The first to 49th frames may be divided into a first divided image and a 50th to 100th frame may be divided into a second divided image.

The motion estimator 120 estimates the background motion and the moving object motion in the divided image. The motion estimation unit 120 determines whether the camera moves in the divided image, and estimates the background motion and the moving body motion based on the movement of the camera. The motion estimation unit 120 includes a camera motion determination unit 122, a background motion estimation unit 124, and a moving object motion estimation unit 126.

The camera motion determiner 122 determines whether a camera motion exists in the divided image. The motion estimation method of the background motion estimator 124 and the moving object motion estimator 126 may be changed based on the determination result of the camera movement.

The camera motion determiner 122 determines the movement of the camera between frames in the divided image by using the brightness of the pixel. For example, the camera motion determiner 122 calculates a ratio of a pixel having a predetermined brightness difference or a pixel having no brightness difference based on the difference image between frames, and the ratio is equal to or less than a preset threshold. If it is determined that the camera is stationary. On the other hand, the camera motion determination unit 122 determines that there is a movement of the camera when the ratio exceeds the preset threshold.

On the other hand, the camera motion determiner 122 may detect the feature points of each frame, and determine the movement of the camera between the frames in the divided image by using the feature point correspondence between the frames. For example, the camera motion discriminating unit 122 calculates a ratio of the movement of the feature point corresponding to the movement of a certain threshold value from the total number of feature point correspondences, and when the ratio of the movement of the feature point correspondence is less than or equal to the predetermined threshold, I think that. On the other hand, when the ratio of the movement amount corresponding to the feature point exceeds a predetermined threshold, the camera motion discriminating unit 122 determines that there is a movement of the camera.

The camera motion determiner 122 may simplify the motion estimation operation of the background motion estimator 124 and the moving object motion estimator 126 when there is no or little camera movement. That is, when it is determined that the camera between frames is in a stationary state, the camera motion discriminating unit 122 estimates the result of the feature point correspondence in the background motion estimating unit 124 and the moving object motion estimating unit 126 as the moving object motion. The area can be treated as being background motion. When the camera between frames is stationary, the camera motion determiner 122 does not need to separately estimate the grid-based background motion in the background motion estimator 124.

The background motion estimator 124 estimates a background motion of the inter-frame static background in the divided image. The background motion estimator 124 may estimate the background motion based on the camera motion determination result of the camera motion determiner 122.

When there is no camera movement between frames in the divided image, the background motion estimator 124 estimates a portion without movement, that is, a pixel region without brightness difference, a feature point portion without movement, and the like as the background motion. In the stationary state without camera movement, the moving object motion estimation unit 126 may estimate the moving part, that is, the pixel area where the brightness difference is greater than or equal to the threshold, the moving feature point part, and the like as the moving object motion.

On the other hand, when there is camera movement between frames in the divided image, the background motion is estimated by the following method.

The background motion estimator 124 divides the frames in the divided image into a grid, estimates the background motion by combining motions of the grids, and estimates the motion of the entire frame. For example, the background motion estimator 124 divides the divided image into predefined grids, and performs regression processing on the homography for each divided grid using the feature point correspondence relationship (F). To estimate the background motion. Here, the feature point correspondence relationship F may extract a feature point by using a scale invariant feature transform (SIFT) method, and analyze the background point by analyzing it using a random SAmple consensus (RANSAC) method. In the case of using the RANSAC method, the background motion estimator 124 estimates the background motion by removing a feature point correspondence whose reprojection error is greater than or equal to a threshold, that is, a feature point correspondence generated from a moving object.

The background motion estimator 124 extracts feature points using the SIFT method and analyzes the background motion by using the RANSAC method. However, the background motion estimator 124 is not necessarily limited thereto. Can be applied. The background motion estimator 124 may estimate the motion of the background using a method of selecting an average motion on a grid basis after matching feature points between frames, or by propagating the motion on a grid basis and applying predetermined spatial filtering. For example, the background motion estimator 124 divides an image into a predefined grid and then propagates the feature point correspondence to the vertices of each grid and performs two-dimensional filtering on each vertex and the image space. The background motion from which the moving object is removed can be estimated.

The moving body motion estimation unit 126 estimates the moving body motion of the frame included in the divided image after estimating the background motion. In detail, the moving object estimator 126 estimates the motion M of the moving object using the difference between the feature point correspondence information F and the background motion B generated based on the feature point correspondence. Here, the moving body motion estimation unit 126 may estimate the moving body motion in the same grid method as that of the background motion estimation. The moving body motion estimation unit 126 may estimate the moving body motion by using Equation 1.

The moving body motion estimation unit 126 may propagate the difference between the feature point correspondence and the background motion to the surroundings or estimate the moving object motion using an average value of the difference between the feature point correspondence and the background motion at a specific position. For example, the moving object motion estimator 126 performs 2D filtering by propagating the difference between the feature point correspondence and the background motion to the peripheral vertices of the grid in order to estimate the grid-based moving object motion, or calculates the average motion around the grid vertices. The motion can be estimated by calculating. Here, the moving body motion estimation unit 126 may apply a penalty function to the difference between the feature point correspondence and the background motion to suppress the minute difference and highlight the large difference to estimate the moving body motion.

The moving body motion estimation unit 126 may extract a moving body motion that is greater than or equal to a predetermined threshold among the moving body motions, and set a grid including the moving body motion as the moving body region.

The image noise remover 130 removes the noise of the image based on the estimated background motion and the moving object motion. The image noise remover 130 includes an image matcher 132, a profile estimator 134, and a filter 136.

The image matching unit 132 sets a window including a plurality of frames in the divided image, and matches the reference frame based on the background motion of the plurality of frames included in the window. In detail, the image matching unit 132 sets a reference frame in a sliding window manner in the divided image, and coordinates a plurality of frames included in the window using the estimated background motion. Sort by.

The image matching unit 132 includes n previous frames t-1, t-2, ..., tn and n subsequent frames t + 1, t + 2,... Based on the reference frame t. , t + n) to match the reference frame t. Here, the matching is preferably processed based on Galilean Transformations, but is not necessarily limited thereto. In the case of n subsequent frames, the image matching unit 132 preferably inversely converts the background motion to match the reference frame t. The background motion up to a specific position may be calculated by summing the inter-frame motion.

Since the image matching unit 132 cannot accurately distinguish the region for the moving object, the image matching unit 132 processes the image registration of the reference frame using only the background motion between the frames.

The profile estimator 134 classifies the moving object region for the moving object in the frame included in the window, and estimates the flow profile for the moving body motion and the background motion because the moving object and the background are mixed in the moving area. . Here, the flow profile refers to information obtained by collecting brightness values over time of a specific pixel position in frames in a window.

As the profile estimator 134 estimates both the flow profile of the moving object and the background in the moving object area, the filtering unit 136 may perform image noise removal by simultaneously considering the flow profile of the moving object and the background motion. On the other hand, the profile estimator 134 estimates only the flow profile of the background motion in the region in which only the background motion exists in the reference frame.

The profile estimator 134 may remove noise in the flow profile of the moving object or the background. Specifically, the profile estimator 134 assumes a distribution of values for color, brightness, etc. included in the flow profile as a normal distribution, removes outliers that deviate from a predetermined standard deviation coefficient, and excludes the removed values. The flow profile can be generated by estimating the mean and variance of the remaining values. For example, the profile estimator 134 assumes that the distribution of the values included in the flow profile is a Gaussian distribution, removes values outside a predetermined sigma coefficient (eg, 3 sigma), and averages the remaining values. And estimate variance to generate a flow profile.

The profile estimator 134 may remove noise in the flow profile by assuming normal distribution for each of the flow profiles of the moving object and the background.

The filtering unit 136 filters the flow profile of the moving object or the background, and removes image noise through the filtering process.

When the filtering unit 136 removes the image noise using only the background flow profile of the background area, the filtering unit 136 performs filtering (1d filtering) on the background flow profile because it is assumed that each pixel corresponding to the background has the same physical location. Can be applied to remove image noise.

On the other hand, when the background area and the moving object area are mixed, the filtering unit 136 compares each of the background flow profile and the moving object flow profile with the reference frame, calculates similarity, and filters the flow profile with high similarity (1d filtering). Eliminate video noise.

Conventionally, noise is removed using only a background motion in a reference frame, and a small area of the moving object is not a problem because the change in color, brightness, etc. is not large. However, when the moving object is large or moves fast, accurate noise removal is impossible. Therefore, the filtering unit 136 according to the present exemplary embodiment may efficiently remove the image noise according to the similarity by comparing each of the background flow profile and the moving object flow profile with the reference frame.

The filtering unit 136 according to the present embodiment compares the average value of each of the background flow profile and the moving object flow profile with the pixel values of the reference frame, selects a flow profile having a small difference, that is, has a high similarity, and filters the selected flow profile. Eliminate video noise. Here, the filtering may be applied to a Gaussian filter, a bilateral filter, a median filter, etc., but the present invention is not limited thereto. Any filter that can be used for image blurring or smoothing may be changed to various types of filters. Applicable

The filtering unit 136 calculates a background similarity between the background flow profile and the pixels of the reference frame using Equation 2 in the moving object region, and moves the moving object between the moving object flow profile and the pixels of the reference frame using Equation 3 Calculate the similarity.

: 기준 프레임 인덱스, t: 분할 영상 인덱스(

-W ≤ t ≤

+W), P_B: 배경 플로우 프로파일, P_M: 이동체 플로우 프로파일, P_B ^E: Edge 영상에서 배경의 edgeness 프로파일 및 P_M ^E: Edge 영상에서 이동체의 edgeness 프로파일)(S _B : background similarity, S _M : mobile similarity, x: pixel position, W: window size,

: Reference frame index, t: Split image index (

-W ≤ t ≤

+ W), P _B : background flow profile, P _M : mobile flow profile, P _B ^E : edgeness profile of background in Edge image and P _M ^E : edgeness profile of moving object in Edge image)

The filtering unit 136 calculates a background similarity and a moving object similarity using Equations 2 and 3, and selects a flow profile having a high similarity among the two similarities. The filtering unit 136 applies filtering (1d filtering) to the selected flow profile to generate an image noise removing result from removing the image noise. Here, the filtering unit 136 is described as applying filtering to the selected flow profile, but is not necessarily limited thereto. Instead, the filtering unit 136 may remove image noise by applying an average value of values in the flow profile. Here, the filtering unit 136 determines that the distance between the reference frame is far, the color is different, and the edgeness difference is large, the similarity is small. The closer the temporal distance to the reference frame is and the color is similar, The smaller the difference in edgeness, the higher the similarity.

2 is a flowchart illustrating a method for removing image noise according to an exemplary embodiment of the present invention.

The image noise canceling apparatus 100 obtains an input image photographed using a camera (S210). The input image may be a video image captured by a camera and may be an image captured by a plurality of spatial regions.

The image noise removing apparatus 100 extracts scene change information of the input image (S212), and generates a split image by dividing the input image by the same spatial region using the scene change information (S220). Here, the scene change information includes a color histogram, accuracy of feature point correspondence, and the like. The image noise removing apparatus 100 generates a segmented image by selecting a frame having a sudden change based on the scene change information as a frame to which the scene is changed. .

The image noise removing apparatus 100 determines the movement of the camera between frames in the divided image (S230). The image noise removing apparatus 100 detects a feature point of each frame and determines the movement of the camera between frames in the divided image using the feature point correspondence between the frames, or is defined in the total number of pixels based on the difference image between the frames. The movement of the camera can be determined by calculating the ratio of pixels having brightness differences or pixels having no brightness differences.

When there is no camera movement in step S230, the image noise removing apparatus 100 initializes a mesh flow of the background frame-to-frame movement to estimate the background motion (S232), and moves the remaining region with the moving object. It is estimated by the motion (S250). Here, the mesh flow for the motion of the background refers to a set of grid-shaped two-dimensional vectors for a part without motion, that is, a pixel area without brightness difference, a feature point part without motion, and the like.

On the other hand, when there is a camera movement in step S230, the image noise removing apparatus 100 divides the frame in the divided image into a grid, and estimates the background motion by combining the grid-specific motions to estimate the motion of the entire frame. In operation S240, the motion of the moving object is estimated using the difference between the feature point correspondence information and the background motion generated based on the feature point correspondence relationship (S250).

The image noise removing apparatus 100 performs an operation of estimating a background motion and a moving object motion in chronological order with respect to a plurality of frames included in the divided image, and when the motion estimation of all the frames in the divided image is completed (S252), the divided image In step S260, a predetermined window for the window sliding process is set.

The image noise removing apparatus 100 matches the reference frame based on the background motion (S270). In detail, the image noise removing apparatus 100 sets a reference frame in a sliding window manner in the divided image, and coordinates a plurality of frames included in the window using the estimated background motion. Sort by ().

The image noise canceller 100 estimates a flow profile of the moving object motion and the background motion, respectively (S280). Specifically, the image noise canceller 100 performs the motion of the moving object in a frame included in the window. Since the moving body region is divided and the moving body and the background are mixed in the moving body region, flow profiles for the moving body motion and the background motion are estimated. Here, the flow profile refers to information obtained by collecting brightness values over time of a specific pixel position in frames in a window.

The image noise removing apparatus 100 filters the flow profile to remove image noise (S290), and generates an image noise removal result (S292). In detail, when the image noise removing apparatus 100 performs the image noise removal on the background region, the image noise removing apparatus 100 removes the image noise by applying filtering (1d filtering) to the background flow profile.

On the other hand, when the image noise removing apparatus 100 performs the image noise reduction for the moving object region, the background flow profile and the moving object flow profile are respectively compared with the pixel values of the reference frame, and the similarity is calculated. Filtering removes image noise.

In FIG. 2, each process is described as being sequentially executed, but is not necessarily limited thereto. In other words, since the process described in FIG. 2 may be applied by changing or executing one or more processes in parallel, FIG. 2 is not limited to the time series order.

The image noise removing method according to the present embodiment described in FIG. 2 may be implemented in an application (or a program) and recorded on a recording medium readable by a terminal device (or a computer). An application (or program) for implementing an image noise canceling method according to the present embodiment is recorded and a recording medium readable by a terminal device (or computer) is any type of recording device in which data that can be read by a computing system is stored. It includes.

3 is an exemplary diagram for describing an operation of removing image noise according to an exemplary embodiment of the present invention.

The image noise removing apparatus 100 classifies and processes the divided image into the divided image 310 of the same space based on the obtained scene change information of the input image.

The image noise removing apparatus 100 estimates the motion 320 for each of the plurality of frames included in the divided image 310. The image noise removing apparatus 100 divides each frame in the divided image 310 into a grid and estimates the grid-based background motion and the moving body motion, respectively.

The image noise canceller 100 performs image registration using a background motion of frames included in a window of the divided image 310. That is, the image noise removing apparatus 100 sets a reference frame by using the background motion, and generates a flow profile for each of the background motion and the moving object motion based on the reference frame.

The image noise removing apparatus 100 filters the one or both of the background flow profile and the moving object profile to remove the image noise (330).

If only the background flow profile exists, the image noise removing apparatus 100 filters the background flow profile to remove the image noise. However, when the moving object flow profile exists, the moving object and the moving object are mixed in the moving area so that the background flow profile and the moving flow profile are compared with the pixels of the reference frame, respectively, to calculate similarity, and to filter the flow profile having high similarity. Eliminate video noise.

4 is an exemplary view for explaining an operation of estimating a motion of a background and a moving object in an image noise removing device according to an exemplary embodiment of the present invention.

The left image of FIG. 4A shows a result of a feature point correspondence relationship between frames in the divided image, and the right image of FIG. 4A shows a result of estimating background motion on a grid basis. The image noise removing apparatus 100 may estimate the background motion by dividing a frame in the divided image into a grid, and combining motions for each grid to estimate motion of the entire frame. For example, the image noise removing device 100 removes a moving object by dividing a frame into grids, then propagating a feature point correspondence to vertices of each grid and performing two-dimensional filtering on each vertex and image space. The background motion can be estimated.

The left image of (b) of FIG. 4 represents an image of the divided image, and the right image of (b) of FIG. 4 represents a result of estimating the moving body motion on a grid basis. The image noise removing apparatus 100 divides a frame in the divided image into a grid, propagates the difference between the feature point correspondence and the background motion in each grid, or uses the average value of the difference between the feature point correspondence and the background motion at a specific position. It can be estimated. For example, the image noise canceller 100 performs 2D filtering by propagating the difference between the feature point correspondence and the background motion to the peripheral vertices of the grid to estimate the grid-based moving object motion, or the motion average value around the vertices of the grid. The motion can be estimated by calculating.

5 is an exemplary diagram for describing an operation of matching an image and removing image noise in an image noise removing apparatus according to an exemplary embodiment of the present invention.

5A is an exemplary diagram of setting a reference frame by performing image registration using a grid-based background motion. As shown in (a) of FIG. 5, the image noise removing apparatus 100 may include one of five frames t-2, t-1, t, t + 1, and t + 2 included in a preset window size. The background motion of some frames t-2, t-1, t + 1, t + 2 is matched with the reference frame t.

FIG. 5B is an exemplary diagram illustrating an operation of selecting a flow profile based on a similarity between the background flow profile and the moving flow profile in order to remove image noise in the moving object region. As shown in (b) of FIG. 5, the image noise removing apparatus 100 extracts a moving object region from a frame included in a window, and compares the background flow profile and the moving body flow profile of the moving object region with pixels of a reference frame. Similarity is calculated. The image noise removing apparatus 100 may remove image noise by selecting a flow profile having a high similarity, that is, a flow profile having a minimum error among the background flow profile and the moving object flow profile.

The above description is merely illustrative of the technical spirit of the embodiments of the present invention, and those skilled in the art to which the embodiments of the present invention pertain various modifications without departing from the essential characteristics of the embodiments of the present invention. Modifications may be possible. Therefore, the embodiments of the present invention are not intended to limit the technical spirit of the embodiments of the present invention, but to describe, and the scope of the technical spirit of the embodiments of the present invention is not limited by these embodiments. The protection scope of the embodiments of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the embodiments of the present invention.

As described above, the embodiment of the present invention is applied to the field of removing image noise, so that a high quality image noise reduction result can be obtained regardless of the motion of the camera sensor and the pattern of the image noise. It is a useful invention to generate an effect that can shorten the time to remove the video noise by removing the video noise by using the image information.

100: video noise reduction device
110: input image processor 120: motion estimation unit
122: camera motion discrimination unit 124: background motion estimation unit
126: moving object motion estimation unit
130: image noise removing unit 132: image matching unit
134: profile estimation unit 136: filtering unit

Claims (9)

In the device for removing image noise in consideration of a moving object,
An input image processor extracting scene change information of an input image and generating a divided image by dividing an image for each scene based on the scene change information;
A motion estimator for checking a movement of a camera in the divided image and estimating a background motion and a moving body motion based on the movement of the camera; And
An image noise remover for removing image noise in an area in which a moving object exists by using the flow profile of the background motion and the moving object motion.
Image noise reduction device comprising a. The method of claim 1,
The image noise removing unit,
An image matching unit which matches a reference frame using the background motion;
A profile estimator for estimating a flow profile of each of the background motion and the moving object motion; And
A filtering unit for removing image noise by comparing each of the background flow profile and the moving body flow profile with the pixel value of the reference image in the region where the moving object exists.
Image noise reduction device comprising a. The method of claim 2,
The filtering unit,
Comparing each of the background flow profile and the moving object flow profile with a pixel value of the reference image to calculate a similarity, and selecting and filtering a flow profile having a high similarity among the background flow profile and the moving object flow profile. Video noise canceller characterized in that. The method of claim 1,
The motion estimation unit,
And detecting a movement of the camera by comparing a feature point corresponding ratio with respect to a difference in brightness or movement of a plurality of pixels among the frames included in the divided image to a preset threshold. The method of claim 4, wherein
The motion estimation unit,
When there is no movement of the camera, the image noise canceller, characterized by estimating a pixel having no brightness difference or a portion of a feature point having no motion as the background motion and estimating a pixel having the brightness difference as the moving body motion. . The method of claim 4, wherein
The motion estimation unit,
When there is motion of the camera, the frames in the divided image are divided into grids, the motions for each grid are collected, the background motions are estimated, and the feature point correspondence information between the frames and the difference between the background motions are used. Image noise canceller, characterized in that the motion estimation. In the method of removing image noise in consideration of a moving object,
An input image processing process of extracting scene change information of an input image and generating a divided image by dividing an image for each scene based on the scene change information;
A motion estimation process of checking a movement of a camera in the divided image and estimating a background motion and a moving body motion based on the movement of the camera; And
An image noise removing process of removing image noise in an area in which a moving object exists by using a flow profile of the background motion and the moving object motion
Image noise reduction method comprising a. The method of claim 7, wherein
The image noise removal process,
An image registration process of matching a reference frame using the background motion;
A profile estimating process of estimating a flow profile for each of the background motion and the moving object motion; And
A filtering process for removing image noise by comparing the background flow profile and the mobile flow profile with the pixel values of the reference image in the region where the moving object exists.
Image noise reduction method comprising a. The method of claim 8,
The filtering process,
Comparing each of the background flow profile and the moving object flow profile with a pixel value of the reference image to calculate a similarity, and selecting and filtering a flow profile having a high similarity among the background flow profile and the moving object flow profile. A video noise reduction method characterized by the above-mentioned.

Images