KR20190072027A - Method for improving low illumination image - Google Patents

Abstract

According to an embodiment of the present invention, a method for improving a low light image comprises the steps of: converting an RGB color space of an input image into an HSV color space; estimating a lighting component and a reflecting component in a V channel of the HSV color space; estimating a weight map to improve the brightness of the reflecting component and to remove a noise; estimating an improved image by using the lighting component, the reflecting component, and the weight map; and combining the improved image and the input image into the RGB color space so as to generate a final image.

Description

[0001] The present invention relates to a low-

The present invention relates to a low-illuminance image enhancement method, and more particularly, to a technique capable of eliminating amplified noise in a process of improving brightness of an image and improving brightness of an image.

Human visual system (HVS) in real world of automobile can deal with dynamic band higher than general image device. Therefore, the conventional image device can not directly capture or display a human-recognized image. Thus, there is a growing need for the processing of images having a dynamic bandwidth larger than that of conventional imaging apparatuses, and these images are called high dynamic range (HDR) images.

The most representative technique for generating a high dynamic range image is to synthesize a plurality of low dynamic range (LDR) images obtained by successively photographing the same scene under different exposure conditions to expand the dynamic range. Specifically, a high dynamic range image can be obtained by synthesizing Short Exposure Image (SEI) and Long Exposure Image (LEI).

In this case, it may take some time to acquire the single exposure image and the long exposure image. In addition, since a single exposure image and a long exposure image are not the same when a subject or an imaging device moves during sequential shooting of a plurality of LDR images, ghost effect in which images overlap in a region where motion occurs during high dynamic range image synthesis .

Also, removing noise from a digital image is very important for acquiring high quality image data. Various noise such as white noise and impulse noise may exist in the digital image data. These noises are added to digital data when a digital image is generated by a digital camera or the like.

The simplest way to remove noise is to use a median filter and a mean filter. Using these filters can remove much of the noise, but blurring can occur where the edge of the image is blurred.

That is, the color information of the high frequency component is distorted while passing through the filter in the portion where the image is rapidly changed. Therefore, edge preserving should be considered in the process of removing noise.

[Patent Literature] Korean Registered Patent No. 1287521.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a low-illuminance image enhancement method that improves brightness according to a bright region and a dark region of an image and effectively removes amplified noise according to an edge region and a flat region of the image.

The technical problems of the present invention are not limited to the above-mentioned technical problems, and other technical problems which are not mentioned can be understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a low-illuminance image enhancement method including transforming an RGB color space of an input image into an HSV color space, estimating an illumination component and a reflection component in a V-channel of the HSV color space, Estimating a weight map for improving brightness and removing noise, estimating an improved image using the illumination component, the reflection component, and the weight map, and combining the improved image and the input image Into an RGB color space to generate a final image.

In one embodiment, the step of estimating the illumination component and the reflection component may estimate the illumination component and the reflection component using the following equations (1) and (2).

[Equation 1]

&Quot; (2) "

(Where f _L is the illumination component, f _R is the reflection component, g _v is the V channel, and H is the Gaussian low-pass filter (mean filter).)

In one embodiment, estimating the weight map may include estimating an inversely transformed brightness channel and estimating an activity map.

In one embodiment, the step of estimating the inversely transformed brightness channel may estimate the inversely transformed brightness channel using the following equation.

[Mathematical Expression]

, 입력영상(g)에서 x에 위치한 R, G, B 픽셀의 화소값 중 가장 큰 값이 위치 x에서의 밝기 채널의 값을 의미한다.)(Where WB (x) is the inverse transformed brightness channel,

, The largest value among the pixel values of the R, G, and B pixels located at x in the input image (g) means the value of the brightness channel at the position x).

In one embodiment, the step of estimating the activity map may estimate the activity map using the following equation.

[Mathematical Expression]

는 x에 위치한 각 화소에서의 표준편차를 의미한다.)(Where A (x) is the active map, H is the Gaussian low-pass filter (mean filter)

Denotes the standard deviation at each pixel located at x.)

In one embodiment, estimating the weight map may estimate the weight map by multiplying the inversely transformed brightness channel by the estimated active map.

[Mathematical Expression]

는 각각의 화소의 위치 x에서 화소끼리의 곱을 의미한다.)(here,

Denotes the product of the pixels at the position x of each pixel.)

In one embodiment, the step of estimating the improved image may improve the brightness of the image using a repetitive slope descent method and remove noise in the edge region.

In one embodiment, the step of converting to the RGB color space may combine the enhanced image and the H channel and the S channel of the input image into an RGB color space.

Calculating a quantization weight value between the step of estimating the weight map and the step of estimating the improved image, the step of improving the brightness of an image using a fast Fourier transform method, And generating a binarized image to synthesize the image.

In one embodiment, the step of calculating the quantization weight value may calculate the quantization weight value using the following equation.

[Mathematical Expression]

(Where W ^q is a quantization weight value, N is a level, W is = 0, 0.33, 0.66, 1 when N is 4, and ^{q is} 1 to 4.)

In one embodiment, the step of improving brightness using the fast Fourier transform method and removing noise in an edge region may be performed using the quantization weight values using Equations 1 and 2: < EMI ID = And generating a reflection component.

[Equation 1]

&Quot; (2) "

은 N개의 반사영상을 의미한다.)(Where W ^q is a quantized weight value,

Denotes N reflection images).

In one embodiment, the step of generating the binarized image comprises: generating a binarization map for synthesizing images according to a quantized q-th weighted level of N reflected components estimated using Equation (1) And generating the binarized image by multiplying the binarization map estimated using Equation (2) by the pixel unit of the estimated N reflection components.

[Equation 1]

&Quot; (2) "

According to the low-illuminance image improving method according to an embodiment of the present invention, weights are applied differently according to a bright region and a dark region in an inversely transformed brightness channel of an image, the weights are applied differently according to the edge region and the flat region to maintain the details of the bright region by using the low weight in the bright region and the edge region by the weight map and minimize blurring in the edge region The noise can be removed, and the weighted map can effectively remove the amplified noise by using the high weight in the dark region and the flat region.

1 is a block diagram schematically illustrating a low-illuminance image enhancement system according to an embodiment of the present invention.
FIG. 2 is a flowchart illustrating a method of improving a low-illuminance image according to an exemplary embodiment of the present invention. Referring to FIG.
3 is a diagram for explaining a method for estimating illumination components and reflection components in a low-illuminance image improving method according to an embodiment of the present invention.
4 is a view for explaining a method of estimating an inversely converted brightness channel in a low-illuminance image improving method according to an embodiment of the present invention.
5 is a diagram illustrating a method of estimating an active map in a low-illuminance image improving method according to an embodiment of the present invention.
6 and 7 are views for explaining a method of estimating a weight map according to an embodiment of the present invention.
FIG. 8 is a flowchart illustrating a fast Fourier transform method and an image enhancement method using a quantization weight value in a low-illuminance image enhancement method according to another embodiment of the present invention.
9 is a diagram for explaining a method of improving brightness and removing noise using a fast Fourier transform method in a low-illuminance image improving method according to an embodiment of the present invention.
10 is a view for explaining a method of generating a binarized image for image synthesis in a low-illuminance image improving method according to an embodiment of the present invention.
11 is a diagram illustrating a computing system that implements a low-illuminance image enhancement method according to an embodiment of the present invention.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference symbols as possible even if they are shown in different drawings. In the following description of the embodiments of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the difference that the embodiments of the present invention are not conclusive.

In describing the components of the embodiment of the present invention, terms such as first, second, A, B, (a), and (b) may be used. These terms are intended to distinguish the constituent elements from other constituent elements, and the terms do not limit the nature, order or order of the constituent elements. Also, unless otherwise defined, all terms used herein, including technical or 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 relevant art and are to be interpreted in an ideal or overly formal sense unless explicitly defined in the present application Do not.

1 is a block diagram schematically illustrating a low-illuminance image enhancement system according to an embodiment of the present invention.

Referring to FIG. 1, the low-illuminance image enhancement system 10 includes an image acquisition unit 100, an image processing unit 200, and an image display unit 300.

The image capturing unit 100 captures an image using a camera. Here, the front camera is described as an embodiment. For example, a camera is an imaging device capable of photographing the front, rear, left, and right sides of a vehicle. For example, the camera shoots up to 8 faces using three or four fisheye lenses, and if the fisheye lens is used, the eight faces can be photographed by the curved evenly.

The image processing unit 200 converts the RGB color space of the input image into the HSV color space, converts the HSV color space to the RGB color space, and converts the RGB color space of the input image and the HSV color space in detail Explain.

The image processing unit 200 estimates the illumination component and the reflection component of the input image, and the method of estimating the illumination component and the reflection component of the input image will be described in detail with reference to FIG.

The image processing unit 200 estimates a weight map in order to improve the brightness of the reflection component of the input image and remove noise, and the method of estimating the weight map will be described in detail with reference to FIG. 6 and FIG.

The image processing unit 200 improves the brightness using a repetitive image enhancement method (repetitive slope descent method), and removes noise in the edge region. For example, in the case of the low-illuminance image enhancement system 10 provided in a vehicle, an image is acquired through the front camera 100, the acquired image is transmitted to the image processing unit 200, Is appropriately processed, and then the image is stored.

That is, the image processing unit 200 improves the image quality by appropriately processing the image before storing the image because it is difficult to distinguish the object from the object due to degraded image quality at night or in an adverse weather condition. An image with improved image quality can be reproduced through the image display unit 300.

FIG. 2 is a flowchart illustrating a method of improving a low-illuminance image according to an exemplary embodiment of the present invention. Referring to FIG.

Referring to FIG. 2, the image processing method in the image processing unit 200 of the low-illuminance image enhancement system 10 according to the embodiment of the present invention will be described in detail with reference to FIG. 1 in the steps S11 to S15 .

In step S11, the image processing unit 200 converts an input image received from the image acquisition unit 100 into an HSV color space from RGB color space (color space) in order to minimize a color distortion occurring in an image processing process, And performs an image processing process on the channel (g _v ). For example, the HSV color space includes H (Hue) channel, S (Saturation) channel and V (luminance) channel, H channel and S channel represent color information of image, ) Or intensity.

In step S12, the image processing unit 200 estimates the illumination component f _L and the reflection component f _R of the input image. For example, the image processing unit 200 may divide the V channel of the input image into an illumination component f _L and a reflection component f _R , and may estimate the illumination component f _L and the reflection component f _R , 3 will be described in detail.

In step S13, the image processing unit 200 estimates a weight map W for improving the brightness with respect to the reflection component f _R and removing the noise. For example, the image processing unit 200 can estimate the inverse-converted brightness channel and the active map, estimate the weight map W using the estimated inverse-converted brightness channel and the active map, The method of estimating the active map will be described in detail with reference to FIG. 4, the method of estimating the active map will be described in detail with reference to FIG. 5, and the method of estimating the weight map W will be described in detail with reference to FIG. 6 and FIG.

)을 추정한다. 예를 들어, 개선된 영상(

)을 추정하는 방법은 하기 수학식 5와 같이 조명성분(f_L), 반사성분(f_R) 및 추정된 가중치 맵(W)이 포함된 에너지 함수를 반복적인 영상 개선 방법(반복적인 기울기 하강 방법)을 이용하여 개선된 영상(

)을 추정할 수 있으며, 개선된 영상(

)을 추정하는 구체적인 방법은 하기 수학식 5 내지 수학식 11을 이용하여 자세하게 설명한다.In step S14, the image processing unit 200 calculates an energy function by using an iterative image enhancement method (an iterative slope descent method)

). For example, improved video (

) Is a method of estimating an energy function including an illumination component f _L , a reflection component f _R and an estimated weight map W as an iterative image enhancement method (an iterative gradient descent method ) To improve the image (

), And the improved image (

) Will be described in detail using the following equations (5) to (11).

)과 입력영상의 H채널 및 S채널을 다시 결합하여 RGB 색상 공간으로 변환하여 최종적인 결과 영상을 생성한다.In step S15, the image processing unit 200 generates an improved image (

) And the H channel and S channel of the input image are recombined and converted into the RGB color space to generate the final result image.

3 is a diagram for explaining a method for estimating illumination components and reflection components in a low-illuminance image improving method according to an embodiment of the present invention.

3 (a), 3 (b), and 3 (c), the input image is a V channel (g _v ) (f _R ), and FIG. 3 (c) is an image obtained by estimating the illumination component (f _L ) in the V channel.

The image processing unit 200 estimates the illumination component f _L and the reflection component f _R from the image using Equation 1 below.

[Equation 1]

Where g _v is the V channel and H is the Gaussian low-pass filter (mean filter).

4 is a view for explaining a method of estimating an inversely converted brightness channel in a low-illuminance image improving method according to an embodiment of the present invention.

4 (a) and 4 (b), FIG. 4 (a) is an image (brightness channel image) obtained by extracting the largest value among RGB pixels in each pixel and FIG. 4 (b) Channel images.

The image processing unit 200 estimates the inversed brightness channel using the RGB image g input by Equation (2).

&Quot; (2) "

이다. 입력영상(g)에서 x에 위치한 R, G, B 픽셀의 화소값 중 가장 큰값(높은값)이 위치 x에서의 밝기 채널의 값이다. 영상 처리부(200)는 0부터 1사이의 값으로 정규화를 실시하고, 1에서 밝기 채널을 차분하여 역변환된 밝기 채널(W_B(x))을 추정한다. Where W _B (x) is the inversely transformed brightness channel,

to be. The largest value (high value) of the pixel values of the R, G, and B pixels located at x in the input image (g) is the value of the brightness channel at the position x. The image processing unit 200 performs normalization with a value between 0 and 1, and estimates an inversely transformed brightness channel W _B (x) by subtracting the brightness channel from 1.

For example, the inversely transformed brightness channel (W _B (x)) is assigned a low weight in the bright region of the image and a large weight is assigned in the dark region of the image, . For example, the image processing unit 200 may apply a low weight to a bright region of an image to weaken the strength of noise elimination, thereby maintaining the detail of the bright region, applying a high weight in the dark region of the image, The amplified noise can be effectively removed when the brightness is improved.

5 is a diagram illustrating a method of estimating an active map in a low-illuminance image improving method according to an embodiment of the present invention.

5 (a) and 5 (b), FIG. 5 (a) is an image (brightness channel image) obtained by extracting the largest value among RGB pixels in each pixel, It is the applied image.

The image processing unit 200 estimates an activation map. Here, the active map is estimated using the following equation (3).

&Quot; (3) "

는 x에 위치한 각 화소에서의 표준편차를 의미한다. Where A (x) is the active map, H is the Gaussian low-pass filter (mean filter)

Is the standard deviation at each pixel located at x.

For example, the image processing unit 200 estimates a low standard deviation in a flat region of the image (excluding an edge region of the image), and estimates a high standard deviation in an edge region of the image.

For example, the image processing unit 200 may apply a high weight to a flat region of an image using a reciprocal of a standard deviation, and apply a low weight to an edge region of the image.

That is, the image processor 200 applies a higher weight to the flat area of the image by effectively removing the amplified noise in improved brightness and, by applying a low weight to the edge of the image edge of the image according to minimize l ₂ -norm It is possible to minimize the blur that occurs in the display device.

6 and 7 are views for explaining a method of estimating a weight map according to an embodiment of the present invention.

6 (a) is an input image, FIG. 6 (b) is an inversely transformed brightness channel image, and FIG. 6 (b) (c) is an image to which an active map is applied, and (d) of FIG. 6 is an image to which a weight map is applied.

The image processing unit 200 estimates a weight map W (x) by multiplying the inversely transformed brightness channel W _B (x) by the active map A (x), as shown in Equation (4) below. The image processing unit 200 can estimate the weight map to improve the brightness of the image and remove noise in the edge region.

&Quot; (4) "

Here, W (x) is a weight map, and x (x) denotes a product of pixels at each pixel position.

Referring to FIGS. 7A and 7B, the image processing unit 200 performs weighting processing corresponding to the bright region A and the dark region B of the image in the inversely transformed brightness channel W _B (x) E and F are applied and the weights G and H are applied to the edge region C and the flat region D of the image in the active map A (x).

For example, the image processing unit 200 may apply a low weight to the bright region and the edge region of the image to weaken the intensity of the noise elimination to maintain the detail of the bright region of the image, the blur can be minimized.

For example, by applying a high weight to dark and flat areas of an image, the image processing unit 200 can enhance the intensity of noise cancellation to effectively remove the amplified noise when the brightness of the dark and flat areas of the image is improved have.

The low-illuminance image enhancement method according to an embodiment of the present invention differs in brightness improvement according to a bright region and a dark region of an image by using an iterative image enhancement method (repetitive slope descent method) It is possible to effectively remove the amplified noise according to the received signal.

For example, the image processing unit 200 can calculate the slope by differentiating the energy function in the image and reduce the slope repeatedly, as shown in Equation (5). In addition, the image processing unit 200 can estimate a solution by minimizing a point at a reduced slope.

The image processing unit 200 can be expressed by the following equations (6) and (7) separately by expressions for the illumination component (f _L ) and the reflection component (f _R )

The image processing unit 200 may differentiate the illumination component f _L and the reflection component f _R in Equations (6) and (7) to estimate solutions as shown in Equations (8) and (9).

The image processing unit 200 calculates a weight map by a repetitive image enhancement method (repetitive gradient descent method) as shown in the following equations (10) and (11) Can be removed. Here, k means the number of times the repetitive image enhancement method is repeated.

&Quot; (5) "

&Quot; (6) "

&Quot; (7) "

&Quot; (8) "

&Quot; (9) "

&Quot; (10) "

&Quot; (11) "

The low-illuminance image enhancement method according to an embodiment of the present invention can speed up the low-level image enhancement method using the fast Fourier transform method and the quantization weight value.

For example, a repetitive image enhancement method (repetitive gradient descent method) has a disadvantage in that the amount of computation is increased and the processing speed is increased due to the repetitive minimization process (repetitive slope minimization process) of the energy function.

Since the image processing unit 200 can not use the spatial domain weight map W _d for the image transformed into the frequency domain by the Fast Fourier Transform (FFT) method in order to improve the processing speed, (W ^q ) obtained by quantizing a weight map having a value of 1 from N to N levels from the quantized weight value as shown in Equation (12) below.

)을 생성하고, 이진화 맵을 이용하여 결과 영상(반사 성분,

)를 적응적으로 합성함으로써, 영상의 밝은 영역 및 어두운 영역에 따른 밝기 개선을 다르게 하고, 영상의 에지 영역 및 평평한 영역에 따른 증폭된 잡음을 효과적으로 제거할 수 있는 결과 영상을 생성할 수 있다.The image processing unit 200 calculates the optimal reflection component f _R by using the weighted values W ^q quantized to the estimated N levels,

), And generates a result image (reflection component,

) Can be adaptively synthesized to improve the brightness according to the bright region and the dark region of the image and to generate the resultant image that effectively removes the amplified noise according to the edge region and the flat region of the image.

For example, the method of increasing the speed using Equation (12) can obtain a result of minimizing the weight of all the pixels of the image using the quantized scalar weight W ^q .

)을 합성한다.The image processor 200 differs in brightness enhancement according to the bright region and the dark region of the image of Equation (5), and generates a quantized weight value < RTI ID = 0.0 > The weight map W (x) is binarized according to the values, and N result images (reflection component,

).

&Quot; (12) "

FIG. 8 is a flowchart illustrating a fast Fourier transform method and an image enhancement method using a quantization weight value in a low-illuminance image enhancement method according to another embodiment of the present invention.

Referring to FIG. 8, in the image processing method in the image processing unit 200 of the low-illuminance image enhancement system 10 according to another embodiment of the present invention, the steps S21 to S28 described above will be described in detail with reference to FIG. 1 .

In step S21, the image processing unit 200 converts an input image received from the image acquisition unit 100 into an HSV color space in an RGB color space (color space) in order to minimize a color distortion occurring in an image processing process, And performs an image processing process on the channel (g _v ). For example, the HSV color space includes H (Hue) channel, S (Saturation) channel and V (luminance) channel, H channel and S channel represent color information of image, ) Or intensity.

In step S22, the image processing unit 200 estimates the illumination component f _L and the reflection component f _R of the input image. For example, the image processing unit 200 can separate the V channel of the input image into the illumination component f _L and the reflection component f _R.

In step S23, the image processing unit 200 estimates a weight map W for improving the brightness with respect to the reflection component f _R and removing the noise. For example, the image processing unit 200 may estimate the inverse-converted brightness channel and the active map, and then estimate the weight map W using the estimated inverse-converted brightness channel and the active map.

In step S24, the image processing unit 200 calculates a quantization weight value. For example, in order to calculate the quantization weight value W ^q , the image processing unit 200 may quantize the weight map having a value of [0 to 1] to N levels as shown in Equation (13).

&Quot; (13) "

W ^q = {0, 0.33, 0.66, 1} and q = 1 to 4 when N is 4 and W ^q is a quantization weight value.

In step S25, the image processing unit 200 improves the brightness using the fast Fourier transform method, makes the brightness of the edge region different from that of the bright region and the dark region of the removed image, A method for improving the brightness using the fast Fourier transform method and eliminating noise in the edge region will be described in detail with reference to FIG.

)을 양자화된 q번째 가중치 레벨에 따라 합성하기 위하여 이진화 맵을 생성한다.In step S26, the image processing unit 200 generates a binarized image for image synthesis, and a method of generating a binarized image will be described in detail with reference to FIG. For example, the image processing unit 200 may estimate the N reflection components (

) According to the quantized q-th weighting level.

)의 픽셀 단위의 곱을 통해 최종적으로 개선된 영상(

)을 획득하고, N개의 개선된 영상(

)을 이진화 맵(B^q(x))을 이용하여 합성함으로써, 영상의 밝기를 개선하고, 밝은 영역의 디테일은 보존하며, 밝기 개선 과정에서 증폭된 잡음을 제거한 결과를 도출할 수 있다.In step S27, the image processing unit 200 synthesizes the images with the improved brightness and noise removed using the binarization map, and synthesizes the images using the binarization map, as described in detail with reference to FIG. For example, the image processing unit 200 may calculate an estimated binarization map B ^q and estimated N reflection components

) Through the product of the pixel in the final image (

), And obtains N improved images (

) Is synthesized by using the binarization map B ^q (x), the brightness of the image is improved, the detail of the bright region is preserved, and the result of removing the amplified noise in the brightness improvement process can be derived.

)과 입력영상의 H채널 및 S채널을 다시 결합하여 RGB 색상 공간으로 변환하여 최종적인 영상을 생성한다.In step S28, the image processing unit 200 receives the improved image (

) And the H and S channels of the input image are recombined and converted into the RGB color space to generate the final image.

9 is a diagram for explaining a method of improving brightness and removing noise using a fast Fourier transform method in a low-illuminance image improving method according to an embodiment of the present invention.

9 (a), 9 (b), 9 (c), and 9 (d), the image processing unit 200 uses the fast Fourier transform method to improve the brightness and remove noise in the edge region, do.

)을 생성하며, 양자화된 가중치 값(W^q,예를 들어, W^q={0, 0.33, 0.66, 1})이 커질수록 잡음 제거의 세기를 강하게 실시할 수 있다.For example, the image processing unit 200 may use the quantized weight values W ^{q to} calculate N reflection components (W ^q ) using the following expressions (14) and (15)

), And the stronger the noise cancellation intensity as the quantized weight value (W ^q , for example, W ^q = {0, 0.33, 0.66, 1}) becomes larger.

&Quot; (14) "

&Quot; (15) "

10 is a view for explaining a method of generating a binarized image for image synthesis in a low-illuminance image improving method according to an embodiment of the present invention.

)을 양자화된 q번째 가중치 레벨에 따라 합성하기 위하여 하기 수학식 16과 같이 이진화 맵을 생성한다.10 (a), (b), (c) and (d), the image processing unit 200 calculates the estimated N reflection components

) According to the quantized q-th weight level, a binary map is generated as shown in Equation (16).

)의 픽셀 단위의 곱을 이용하여 최종적으로 개선된 영상(

)을 획득하고, N개의 개선된 영상(

)을 이진화 맵(B^q(x))을 이용하여 합성함으로써, 밝기를 개선하고, 에지 영역의 잡음을 제거한 영상을 추출할 수 있다.The image processing unit 200 may calculate the estimated binomial B ^q and the estimated N reflection components

), Which is the final improved image (

), And obtains N improved images (

) Is synthesized by using the binarization map B ^q (x), it is possible to improve the brightness and extract the image from which noise in the edge region is removed.

&Quot; (16) "

&Quot; (17) "

11 is a diagram illustrating a computing system that implements a low-illuminance image enhancement method according to an embodiment of the present invention.

11, a computing system 1000 includes at least one processor 1100, a memory 1300, a user interface input device 1400, a user interface output device 1500, (1600), and a network interface (1700).

The processor 1100 may be a central processing unit (CPU) or a memory device 1300 and / or a semiconductor device that performs processing for instructions stored in the storage 1600. Memory 1300 and storage 1600 may include various types of volatile or non-volatile storage media. For example, the memory 1300 may include a ROM (Read Only Memory) and a RAM (Random Access Memory).

Thus, the steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by processor 1100, or in a combination of the two. The software module may reside in a storage medium (i.e., memory 1300 and / or storage 1600) such as a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, You may. An exemplary storage medium is coupled to the processor 1100, which can read information from, and write information to, the storage medium. Alternatively, the storage medium may be integral to the processor 1100. [ The processor and the storage medium may reside within an application specific integrated circuit (ASIC). The ASIC may reside within the user terminal. Alternatively, the processor and the storage medium may reside as discrete components in a user terminal.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention.

Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

10: Low light image enhancement system
100:
200:
300:
1000: Computing System
1100: Processor
1200: System bus
1300: Memory
1310: ROM
1320: RAM
1400: User interface input device
1500: User interface output device
1600: Storage
1700: Network interface

Claims (12)

Converting an RGB color space of an input image into an HSV color space;
Estimating an illumination component and a reflection component in a V-channel of the HSV color space;
Estimating a weight map for improving the brightness of the reflection component and removing noise;
Estimating an improved image using the illumination component, the reflection component, and the weight map; And
Combining the improved image and the input image to convert the input image into an RGB color space to generate a final image
Wherein the low-illuminance image enhancing method comprises: The method according to claim 1,
Wherein the step of estimating the illumination component and the reflection component comprises:
Wherein the illumination component and the reflection component are estimated using the following equations (1) and (2).
[Equation 1]

&Quot; (2) "

(Where f _L is the illumination component, f _R is the reflection component, g _v is the V channel, and H is the Gaussian low-pass filter (mean filter).) The method according to claim 1,
Estimating the weight map comprises:
Estimating an inversely transformed brightness channel; And
And estimating an active map based on the first and second images. The method of claim 3,
The step of estimating the inversely converted brightness channel comprises:
Wherein the inversely transformed brightness channel is estimated using the following equation.
[Mathematical Expression]

(Where W _B (x) is the inverse transformed brightness channel,

, The largest value among the pixel values of the R, G, and B pixels located at x in the input image (g) means the value of the brightness channel at the position x). The method of claim 3,
Wherein estimating the activity map comprises:
Wherein the active map is estimated using the following equation.
[Mathematical Expression]

(Where A (x) is the active map, H is the Gaussian low-pass filter (mean filter)

Denotes the standard deviation at each pixel located at x.) The method of claim 3,
Estimating the weight map comprises:
And estimating the weight map by multiplying the inversely transformed brightness channel by the estimated active map.
[Mathematical Expression]

(here,

Denotes the product of the pixels at the position x of each pixel.) The method according to claim 1,
Wherein the estimating the enhanced image comprises:
Wherein the brightness of the image is improved by using the repetitive slope descent method, and the noise of the edge region is removed. The method according to claim 1,
Wherein the converting into the RGB color space comprises:
Wherein the improved image and the H and S channels of the input image are combined and converted into an RGB color space. The method according to claim 1,
Between the step of estimating the weight map and the step of estimating the improved image,
Calculating a quantization weight value;
Improving the brightness of an image using a fast Fourier transform method and removing noise in an edge region; And
Generating a binarized image to synthesize the image;
Wherein the low-illuminance image enhancing method comprises: The method of claim 9,
Wherein the step of calculating the quantization weight value comprises:
Wherein the quantization weight value is calculated using the following equation.
[Mathematical Expression]

(Where W ^q is a quantization weight value, N is a level, W is = 0, 0.33, 0.66, 1 when N is 4, and ^{q is} 1 to 4.) The method of claim 9,
Wherein the step of improving the brightness using the fast Fourier transform method and removing the noise of the edge region comprises:
And generating a reflection component using the quantization weight value using Equation (1) and Equation (2) below.
[Equation 1]

&Quot; (2) "

(Where W ^q is a quantized weight value,

Denotes N reflection images.) The method of claim 9,
Wherein the generating the binarized image comprises:
Generating a binarization map for synthesizing images according to a quantized q-th weighted level of N reflected components estimated using Equation (1) below; And
And generating the binarized image by multiplying the estimated binarization map using the following equation (2) by the pixel unit of the estimated N reflection components.
[Equation 1]

&Quot; (2) "

Images