CN105005973A - Fast image denoising method and apparatus - Google Patents

Abstract

The embodiment of the present invention discloses a method and device for rapid denoising of an image, wherein the method for rapid denoising of an image includes: dividing the image into regions according to the brightness values of pixels in the image; Region, use the denoising strength corresponding to the current region to denoise the current region. Using the technical solution provided by the embodiment of the present invention, a smaller denoising intensity is used for areas with higher brightness to ensure that image details in areas with higher brightness are not lost, and a larger denoising intensity can be used for areas with lower brightness , can effectively control the noise, maintain the integrity of the original information in the low brightness area, can effectively improve the image quality, increase the signal-to-noise ratio, and better reflect the information carried by the original image.

Description

A method and device for quickly denoising an image

technical field

Embodiments of the present invention relate to the field of image processing, and in particular, to a method and device for fast image denoising.

Background technique

With the popularization of digital products, images and videos have become the most commonly used information carriers in human activities. They contain a lot of information about objects and become the main way for people to obtain original information from the outside world. However, in the process of image acquisition, transmission and storage, it is often interfered and affected by various noises, which degrades the image quality. In order to obtain high-quality digital images, it is necessary to denoise the images to remove useless information in the signal while maintaining the integrity of the original information (that is, the main features) as much as possible.

The currently used denoising algorithms all select different denoising strengths according to the ambient light brightness of the image. Although this algorithm can achieve a very good denoising effect when the ambient brightness of the image is consistent, but for an image with bright and dark areas, if the denoising degree is large, the details of the highlighted area will be wiped out. ; If the degree of denoising is small, it will lead to severe noise in low-brightness areas.

Contents of the invention

In view of this, an embodiment of the present invention proposes a method and device for fast image denoising to solve the problem that image denoising effects are affected by inconsistencies in image background brightness.

In a first aspect, an embodiment of the present invention provides a method for fast image denoising, the method comprising:

performing region division on the image according to the brightness values of the pixels in the image;

For each region, the denoising intensity corresponding to the current region is used to perform denoising processing on the current region.

In the second aspect, an embodiment of the present invention provides a device for fast image denoising, the device comprising:

an area division unit, configured to perform area division on the image according to the brightness values of pixels in the image;

The denoising processing unit is configured to perform denoising processing on the current area for each area using the denoising intensity corresponding to the current area.

Using the technical solution provided by the embodiment of the present invention, the image is denoised by selecting different denoising strengths according to the brightness values of different regions of the image, and a smaller denoising strength is used for regions with higher brightness to ensure that the brightness is lower. The image details in the high-brightness area are not lost, and a larger de-noising intensity can be used for the low-brightness area, which can effectively control the noise and maintain the integrity of the original information in the low-brightness area, which can effectively improve the image quality and increase the signal noise. better reflect the information carried by the original image.

Description of drawings

Other characteristics, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

FIG. 1 is a flowchart of a method for fast image denoising provided in the first embodiment of the present invention;

FIG. 2 is a flowchart of a method for fast image denoising provided by the second embodiment of the present invention;

FIG. 3 is a flow chart of a method for fast image denoising provided by a third embodiment of the present invention;

FIG. 4 is a flowchart of a method for fast image denoising provided by the fourth embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an apparatus for fast image denoising provided by a fifth embodiment of the present invention.

detailed description

The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, but not to limit the present invention. In addition, it should be noted that, for the convenience of description, only parts related to the present invention are shown in the drawings but not all content.

Figure 1 shows a first embodiment of the invention.

Fig. 1 is a schematic flowchart of a method for fast image denoising provided by the first embodiment of the present invention. The method of the embodiment of the present invention can be executed by a fast image denoising device, which can be implemented by means of software and/or hardware Realized, integrated in the smart terminal with camera function.

Referring to Fig. 1, the method for fast image denoising includes:

Step S101 , divide the image into regions according to the brightness values of the pixels in the image.

When using a photographing device, such as a digital camera, a mobile phone or other smart terminals to take pictures, due to the object's occlusion of light, there will be bright and dark areas in the photographed object itself. The bright area and the dark area are formed due to the large difference in the brightness of the image. The brightness of the image refers to the intensity of the image pixel. Black is the darkest and white is the brightest. Generally, black is represented by 0, and white is represented by 255. . By converting the image into HSL space, the brightness value of the pixel in the image can be obtained through the L component value. In addition, during the photographing process, according to the Y component value of the YUV image transmitted by the photographing device, the brightness value of the pixel in the image can also be obtained, and the image can be divided into multiple regions according to the brightness value of the pixel in the image .

Step S102 , for each region, denoise the current region using the denoising intensity corresponding to the current region.

The camera device converts optical signals into digital signals. The captured image is noisy due to electronic interference during the conversion process. In order to get better image quality, it is necessary to denoise the image. In the image, areas with higher brightness have lower noise, and areas with lower brightness have higher noise, so it is necessary to use different denoising intensities for different areas to perform denoising processing on the areas. The denoising strength refers to the removal coefficient of the noise imaging data. According to the regions divided in step S101, denoising processing is performed on the image according to the denoising intensity corresponding to the current region. By using different denoising strengths to remove the noise in the image in the area with large brightness difference, it can remove the image noise while ensuring the fidelity of the image and retaining the image details in the area with less noise.

In the technical solution provided by this embodiment, the image is denoised by selecting different denoising intensities according to the brightness values of different areas of the image, and a lower denoising intensity is used for areas with higher luminance to ensure that the areas with higher luminance The image details in the area are not lost; for areas with low brightness, a higher denoising intensity can be used, which can effectively control noise, effectively improve image quality, increase the signal-to-noise ratio, and better reflect the information carried by the original image .

On the basis of the above technical solution, in this embodiment, before step S101 "dividing the image into regions according to the brightness values of the pixels in the image", the following steps can also be added: receiving a photographing instruction, turning on the camera , using the camera to collect images.

The receiving of the photographing instruction may be that the photographing device receives a preset corresponding operation performed by the user to start the photographing application, for example, the photographing device receives the user's pressing operation on a physical button on the photographing device, or receives the The corresponding tap or swipe action on the smartphone touch screen. When the photographing device receives the photographing instruction, it turns on the camera, and collects external images through the camera.

Figure 2 shows a second embodiment of the invention.

Fig. 2 is a flow chart of a method for fast image denoising provided by the second embodiment of the present invention. The method for rapid image denoising is based on the first embodiment of the present invention. Further, the divided regions are divided into: high-brightness regions and low-brightness regions; The brightness value is used to divide the image into regions, and the specific optimization is as follows: divide the image into multiple regions according to preset rules, and obtain the average brightness value of the pixels in the regions; divide the pixels in the regions The average value of the luminance value is compared with the preset first threshold and the second threshold respectively, and the area is divided into a high-brightness area or a low-brightness area according to the comparison result.

Referring to Fig. 2, the method for fast denoising of the image includes:

In step S201, the image is divided into multiple regions according to preset rules, and the average brightness value of pixels in the regions is obtained.

The image is divided into multiple regions, and the division rules can be preset, for example, the size and orientation of each region can be preset. In this embodiment, the image is divided into 16×16 image areas according to the size of the image, and the luminance values of all pixels in the divided areas are obtained, and the pixels in the area are calculated according to the luminance values. The average brightness value of the point.

Step S202 , comparing the average brightness values of the pixels in the area with preset first thresholds and second thresholds, and classifying the area as a high-brightness area or a low-brightness area according to the comparison result.

The average brightness value of the pixels in the area obtained in step S201 is compared with a preset first threshold and a second threshold, and the first threshold and the second threshold are preset by the system or the user. When the obtained average brightness value of the pixels in the region is greater than the first threshold, the region is classified as a highlighted region. For example, by exposing a YUV image, the average value of the Y value of the pixels in a certain area is obtained. If the average value of the Y value is greater than 200, the area is divided into a highlighted area; the brightness of the pixel points in the acquired area When the average value is smaller than the second threshold, the region is classified as a low-brightness region. For example, by exposing the YUV image, the average Y value of the pixels in a certain area is obtained. If the average Y value is less than 50, the area is divided into a low-brightness area.

As another optional implementation manner of this embodiment, other luminance value characteristics in the region may also be used as comparison conditions to determine whether the region is a high-brightness or low-brightness region. For example, the sum of the brightness values of pixels in the area is used as the brightness value feature. Comparing the acquired sum of brightness values of pixels in the region with preset first brightness value and threshold value and second brightness value and threshold value respectively, classifying the region as a high-brightness region or a low-brightness region.

Step S203, for each region, respectively, adopt the denoising intensity corresponding to the current region to perform denoising processing on the current region.

In this embodiment, the image is divided into regions according to the brightness values of the pixels in the image, and the specific optimization is as follows: the image is divided into multiple regions according to preset rules, and the pixels in the regions are obtained The average brightness value of the pixel points in the area is compared with the preset threshold value, and the area is determined to be a high-brightness or low-brightness area according to the comparison result. The method can accurately divide the high-brightness area and the low-brightness area according to the luminance distribution characteristics of the image, and has the advantages of low calculation amount, quick response, easy implementation and the like compared with other division methods.

Figure 3 shows a third embodiment of the invention.

FIG. 3 is a flow chart of a method for fast image denoising provided by a third embodiment of the present invention. The method for rapid image denoising is based on the first embodiment of the present invention. Further, the denoising process for the current area is optimized specifically for each area using the denoising intensity corresponding to the current area as follows: : Respectively for each area, identify the target pixel points that are noise points in the current area; remove the target pixel points that are noise points in the image to obtain a noise-free image, and use the preset noise reduction algorithm to perform noise reduction on the noise-free image Processing to obtain a new image; multiplying the imaging data of the target pixel with the anti-denoising strength coefficient corresponding to the current area to obtain the anti-denoising strength data corresponding to the target pixel; adding the target pixel to the new image The anti-denoising strength data corresponding to the point.

Referring to Fig. 3, the method for fast denoising of the image includes:

Step S301, divide the image into regions according to the brightness values of the pixels in the image.

Step S302, for each area, identify target pixel points that are noise points in the current area.

Respectively identify the noise points in the area divided by step S301. Because the noise points are generated by electronic interference, they are completely different from the surrounding pixels, and visually look similar to the image being dirty compared with other pixels. Effect. Using this feature, it is possible to identify pixels that are noise points in the area and use these pixels as target pixels.

Step S303 , removing the noise-free target pixels in the image to obtain a noise-free image, and performing noise-reduction processing on the noise-free image using a preset noise-reduction algorithm to obtain a new image.

The target pixel points obtained in step S302 are deleted from the image to obtain a noise-free image, so that the target pixel points of the noise-free image in the original image are empty (null). The noise-free image is denoised according to a preset noise-reduction algorithm, and the noise-reduction method may select known noise-reduction algorithms such as median filtering, Gaussian filtering, or binary filtering to denoise the noise-free image. In this embodiment, a noise-free image may be denoised by using a wavelet transform denoising algorithm. Wavelet transform can find the best approximation to the original signal, so as to complete the distinction between the original signal and the noise signal, so as to obtain the best restoration of the original signal.

In step S304, the imaging data of the target pixel is multiplied by the anti-denoising strength coefficient corresponding to the current region to obtain the anti-denoising strength data corresponding to the target pixel.

The imaging data of the target pixel point is the data stored by the pixel point in the image, for example, the RGB component values of the pixel point in the color image, or the gray value of the pixel point in the image. Exemplarily, the imaging data of the target pixel is selected as the brightness value of the target pixel, and the anti-denoising strength coefficient is preset by the system. The anti-denoising strength coefficient reflects the degree of retention of the noise data in the image. The larger the anti-denoising strength coefficient is, the more noise imaging data is retained, and the lower the denoising strength is; The less imaging data is preserved, the stronger the denoising will be. In this embodiment, the anti-denoising strength coefficient is in the range of 0-1 and can be determined empirically. For different regions, the preset anti-denoising strength coefficients are also different. For example, for the highlighted area, due to its low noise, in order to retain more image details, a higher anti-denoising strength coefficient can be used, for example, 0.8, which can retain more noise imaging data; correspondingly, For low-brightness areas, a lower anti-denoising strength coefficient, such as 0.2, can be used to remove more noise imaging data to achieve a better denoising effect and make the processed image smoother. The result of multiplying the imaging data of the target pixel by the anti-denoising strength coefficient corresponding to the current region is used as the anti-denoising strength data corresponding to the target pixel.

Step S305, adding the anti-denoising strength data corresponding to the target pixel on the new image.

Generally, an image is represented by an m×n pixel matrix. The new image in step S303 is an m×n pixel matrix including part of the pixel imaging data which is empty (null) due to the removal of noise. Add the anti-denoising strength data corresponding to the target pixel on the new image, specifically: replace the anti-denoising strength data in the new image pixel matrix with empty pixel imaging data according to the position of the original noise point Pixels, get a new m×n pixel matrix. The effect produced by the above steps is similar to superimposing the anti-denoising strength data corresponding to the target pixel on the new image without noise after the noise reduction processing according to the original position of the target pixel, so that the new The image is filled with all the pixel data to become a complete image after denoising processing.

In this embodiment, the specific optimization of the denoising process on the current area is carried out by using the denoising intensity corresponding to the current area for each area respectively as follows: for each area, identify the target pixel points that are noise points in the current area; remove The noise-free target pixel in the image is obtained to obtain a noise-free image, and a preset noise reduction algorithm is used to perform noise-reduction processing on the noise-free image to obtain a new image; Perform a product operation on the denoising strength coefficient to obtain the anti-denoising strength data corresponding to the target pixel; add the anti-denoising strength data corresponding to the target pixel to the new image. It can quickly and effectively denoise the image captured by the camera device, and select the appropriate denoising intensity according to the brightness of the image area, so that the denoised image can not only maintain image details in the high-brightness area; The low-brightness area effectively de-noises the noise, making the image smoother.

Fig. 4 shows a fourth embodiment of the present invention.

Fig. 4 is a flow chart of a method for fast image denoising provided by a fourth embodiment of the present invention. The method for rapid denoising of an image is based on the third embodiment of the present invention, and further, to identify target pixels that are noises in the current area, the specific optimization is: to obtain each pixel in the current area and its adjacent pixels The minimum value of the brightness difference of the point, when the minimum value of the brightness difference is greater than the preset minimum value threshold, the pixel point is regarded as a noise point.

Referring to Fig. 4, the method for fast image denoising includes:

Step S401, divide the image into regions according to the brightness values of the pixels in the image.

Step S402, obtaining the minimum value of the brightness difference between each pixel in the current area and its adjacent pixels, and when the minimum brightness difference is greater than a preset minimum value threshold, the pixel is regarded as a noise point.

Noise is generated by electronic interference, and its characteristics are completely different from those of surrounding pixels. The feature may be a brightness value of a pixel point or a color RGB component or the like. In this embodiment, the feature used is the brightness value. The brightness values of adjacent pixels around each pixel in the current area are sequentially acquired, and the brightness difference between each pixel and its adjacent pixels is calculated. By comparing the brightness difference, the minimum value of the brightness difference is obtained. And comparing the minimum value of the brightness difference with a preset minimum value threshold, and when the minimum value of the brightness difference is greater than the preset minimum value threshold, the pixel is regarded as a noise point. Exemplarily, the adjacent pixel points may be 8 adjacent pixels. The luminance values of 8 pixels adjacent to the pixel point are subtracted from the luminance value of the pixel point in turn, and the difference values are compared to determine the minimum value of the 8 luminance value difference values. The minimum value is compared with a preset minimum value threshold, and the minimum value threshold can be set by experience. When the minimum value of the luminance value difference is greater than the preset minimum value threshold, it indicates that the difference between the pixel point and surrounding pixel points exceeds a normal range, and the pixel point is determined to be a noise point.

Step S403 , removing the noise-free target pixels in the image to obtain a noise-free image, and performing noise-reduction processing on the noise-free image using a preset noise-reduction algorithm to obtain a new image.

Step S404, performing a product operation on the imaging data of the target pixel point and the anti-denoising strength coefficient corresponding to the current area to obtain the anti-denoising strength data corresponding to the target pixel point.

Step S405, adding the anti-denoising strength data corresponding to the target pixel on the new image.

In this embodiment, by identifying target pixels that are noises in the current area, the specific optimization is as follows: obtaining the minimum value of the brightness difference between each pixel in the current area and its adjacent pixels, and when the minimum value of the brightness difference is greater than When the preset minimum value threshold is used, the pixel is regarded as a noise point. It can accurately and quickly identify noise points, making the subsequent denoising processing of noise points more effective.

Fig. 5 shows a fifth embodiment of the present invention.

Fig. 5 is a structural diagram of an apparatus for fast image denoising provided by a fourth embodiment of the present invention. Referring to FIG. 5 , the device for rapid image denoising includes: an area division unit 510 and a denoising processing unit 520 .

Wherein, the area division unit 510 is configured to perform area division on the image according to the brightness values of pixels in the image;

The denoising processing unit 520 is configured to perform denoising processing on the current region by using the denoising intensity corresponding to the current region for each region.

In the technical solution provided by this embodiment, the image is denoised by selecting different denoising strengths according to the brightness values of different regions of the image, and a smaller denoising strength is used for regions with higher brightness to ensure that the regions with higher brightness The image details in the area are not lost, and a larger denoising intensity can be used for areas with low brightness. It can effectively control noise, effectively maintain the integrity of original information in areas with low brightness, and can effectively improve image quality and increase signal noise. better reflect the information carried by the original image.

Further, the area division unit 510 includes: a brightness value average acquisition subunit 511 , a high brightness area division subunit 512 and a low brightness area division subunit 513 .

Wherein, the average brightness value acquisition subunit 511 is configured to divide the image into multiple regions according to preset rules, and obtain the average brightness value of pixels in the regions;

The highlight area division subunit 512 is configured to divide the area into a highlight area when the average brightness value of pixels in the area is higher than a preset first threshold;

The low-brightness area division subunit 513 is configured to divide the pixel into a low-brightness area when the average brightness value of the pixels in the area is lower than a preset second threshold.

Further, the denoising processing unit 520 includes: a target pixel identification subunit 521 , a new image acquisition subunit 522 , an anti-denoising strength data acquisition subunit 523 and an anti-denoising strength data addition subunit 524 .

Wherein, the target pixel point identification subunit 521 is used to identify target pixel points that are noise points in the current area for each area;

The new image acquisition subunit 522 is configured to remove target pixels that are noise points in the image to obtain a noise-free image, and use a preset noise reduction algorithm to perform noise reduction processing on the noise-free image to obtain a new image;

The anti-denoising strength data acquisition subunit 523 is used to multiply the imaging data of the target pixel with the anti-denoising strength coefficient corresponding to the current area to obtain the anti-denoising strength data corresponding to the target pixel;

The anti-denoising strength data adding subunit 524 is configured to add the anti-denoising strength data corresponding to the target pixel on the new image.

Further, the target pixel identification subunit 521 is used for:

Acquire the minimum value of the brightness difference between each pixel in the current area and its adjacent pixels, and when the minimum brightness difference is greater than a preset minimum value threshold, the pixel is regarded as a noise point.

Furthermore, the device for rapid image denoising further includes: an image acquisition unit 530,

Wherein, the image collection unit 530 is configured to receive a photographing instruction, turn on the camera, and use the camera to collect images.

The above device for fast image denoising can execute the method for fast image denoising provided in the embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method.

The serial numbers of the above embodiments of the present invention are for description only, and do not represent the advantages and disadvantages of the embodiments.

Those of ordinary skill in the art should understand that each module or each step of the present invention described above can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed on a network formed by multiple computing devices. Optionally, they can be implemented with executable program codes of computer devices, so that they can be stored in storage devices and executed by computing devices, or they can be made into individual integrated circuit modules, or a plurality of modules in them Or the steps are fabricated into a single integrated circuit module to realize. As such, the present invention is not limited to any specific combination of hardware and software.

Each embodiment in this specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same or similar parts between the various embodiments can be referred to each other.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims (10)

1. A method for fast image denoising, comprising: performing region division on the image according to the brightness values of the pixels in the image; For each region, the denoising intensity corresponding to the current region is used to perform denoising processing on the current region. 2. The method according to claim 1, wherein the divided regions include: a high-brightness region and a low-brightness region; According to the brightness value of the pixel in the image, the image is divided into regions, including: dividing the image into a plurality of regions according to preset rules, and acquiring the average brightness value of pixels in the regions; If the average brightness value of the pixels in the area is higher than the preset first threshold, the area is divided into a highlighted area; If the average brightness value of the pixels in the area is lower than the preset second threshold, the pixel is classified as a low-brightness area. 3. The method according to claim 1, wherein the denoising process is performed on the current area by using the denoising intensity corresponding to the current area for each area, including: For each area, identify the target pixel points that are noise in the current area; removing noise-free target pixels in the image to obtain a noise-free image, and performing noise reduction processing on the noise-free image using a preset noise reduction algorithm to obtain a new image; Performing a product operation on the imaging data of the target pixel and the anti-denoising strength coefficient corresponding to the current region to obtain the anti-denoising strength data corresponding to the target pixel; Add the anti-denoising strength data corresponding to the target pixel on the new image. 4. The method according to claim 3, wherein identifying target pixels that are noise points in the current region comprises: Acquire the minimum value of the brightness difference between each pixel in the current area and its adjacent pixels, and when the minimum brightness difference is greater than a preset minimum value threshold, the pixel is regarded as a noise point. 5. The method according to any one of claims 1-4, wherein, before performing region division on the image according to the brightness values of pixels in the image, further comprising: After receiving the photographing instruction, the camera is turned on, and the camera is used to collect images. 6. A device for fast denoising of an image, comprising: an area division unit, configured to perform area division on the image according to the brightness values of pixels in the image; The denoising processing unit is configured to perform denoising processing on the current area for each area using the denoising intensity corresponding to the current area. 7. The device according to claim 6, wherein the area dividing unit comprises: The average brightness value acquisition subunit is used to divide the image into multiple regions according to preset rules, and obtain the average brightness value of pixels in the regions; A highlight area division subunit, configured to divide the area into a highlight area when the average brightness value of the pixels in the area is higher than a preset first threshold; The low-brightness area division subunit is configured to divide the pixel into a low-brightness area when the average brightness value of the pixels in the area is lower than a preset second threshold. 8. The device according to claim 6, wherein the denoising processing unit comprises: The target pixel point identification subunit is used to identify target pixel points that are noise points in the current area for each area; A new image acquisition subunit, configured to remove target pixels that are noise points in the image to obtain a noise-free image, and use a preset noise reduction algorithm to perform noise reduction processing on the noise-free image to obtain a new image; The anti-denoising strength data acquisition subunit is used to perform a product operation on the imaging data of the target pixel and the anti-denoising strength coefficient corresponding to the current area, to obtain the anti-denoising strength data corresponding to the target pixel; The anti-denoising intensity data adding subunit is configured to add the denoising intensity data corresponding to the target pixel on the new image. 9. The device according to claim 7, wherein the denoising intensity data obtains a subunit for: Acquire the minimum value of the brightness difference between each pixel in the current area and its adjacent pixels, and when the minimum brightness difference is greater than a preset minimum value threshold, the pixel is regarded as a noise point. 10. The device according to claim 6, further comprising: The image acquisition unit is configured to receive a photographing instruction, turn on the camera, and use the camera to collect images.