WO2020103503A1 - 夜景图像处理方法、装置、电子设备及存储介质 - Google Patents

夜景图像处理方法、装置、电子设备及存储介质

Info

Publication number
WO2020103503A1
WO2020103503A1 PCT/CN2019/101430 CN2019101430W WO2020103503A1 WO 2020103503 A1 WO2020103503 A1 WO 2020103503A1 CN 2019101430 W CN2019101430 W CN 2019101430W WO 2020103503 A1 WO2020103503 A1 WO 2020103503A1
Authority
WO
WIPO (PCT)
Prior art keywords
noise reduction
target area
image
preset
target
Prior art date
Application number
PCT/CN2019/101430
Other languages
English (en)
French (fr)
Inventor
黄杰文
Original Assignee
Oppo广东移动通信有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Oppo广东移动通信有限公司 filed Critical Oppo广东移动通信有限公司
Publication of WO2020103503A1 publication Critical patent/WO2020103503A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/80—Camera processing pipelines; Components thereof
    • H04N23/81—Camera processing pipelines; Components thereof for suppressing or minimising disturbance in the image signal generation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/70—Circuitry for compensating brightness variation in the scene
    • H04N23/73—Circuitry for compensating brightness variation in the scene by influencing the exposure time
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/70—Circuitry for compensating brightness variation in the scene
    • H04N23/741—Circuitry for compensating brightness variation in the scene by increasing the dynamic range of the image compared to the dynamic range of the electronic image sensors

Definitions

  • the present application relates to the field of imaging technology, and in particular, to a night scene image processing method, device, electronic device, and storage medium.
  • Both mobile phones and tablet computers have built-in cameras, and with the enhancement of mobile terminal processing capabilities and the development of camera technology, the performance of the built-in camera is getting stronger and stronger, and the quality of the captured images is getting higher and higher.
  • the operation of mobile terminals is now simple and easy to carry. In daily life, people have taken pictures with mobile terminals such as smartphones and tablet computers, which has become a normal state.
  • the night scene image processing method, device, electronic equipment and storage medium proposed in this application are used to solve the problem of using the traditional noise reduction method to reduce the noise of the night scene captured images in the related art, which cannot guarantee the noise reduction effect of the sky area and other The detailed information of the area is kept good, affecting the user experience.
  • the night scene image processing method proposed by an embodiment of the present application on the one hand includes: sequentially acquiring multiple frames of images according to a preset exposure compensation mode; and using a preset image recognition model to perform on the preset frames of the multiple frames of images Identification processing to determine the target area and non-target area in the preset frame image; using different noise reduction parameter values, the target area and the non-target area of the multi-frame image are respectively subjected to noise reduction processing to generate a target In the image, the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the non-target area.
  • the night scene image processing device includes: an acquisition module for sequentially acquiring multiple frames of images according to a preset exposure compensation mode; a determination module for using a preset image recognition model Performing identification processing on the preset frame image in the multi-frame image to determine the target area and the non-target area in the preset frame image; a noise reduction module, configured to use different noise reduction parameter values
  • the target area and the non-target area of the image are respectively subjected to noise reduction processing to generate a target image, wherein the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the non-target area.
  • An electronic device provided in an embodiment of another aspect of the present application includes the camera module, memory, processor, and a computer program stored on the memory and executable on the processor, characterized in that the processor executes The program implements the night scene image processing method described above.
  • a computer-readable storage medium provided in an embodiment of another aspect of the present application, on which a computer program is stored, and is characterized in that the program for night scene image processing described above is implemented when the program is executed by a processor.
  • a computer program proposed by another embodiment of the present application When the program is executed by a processor, the night scene image processing method described in the embodiment of the present application is implemented.
  • the night scene image processing method, device, electronic device, computer-readable storage medium, and computer program provided in the embodiments of the present application can sequentially acquire multiple frames of images according to a preset exposure compensation mode, and use the preset image recognition model to Recognize the preset frame image in the multi-frame image to determine the target area and non-target area in the preset frame image, and then use different noise reduction parameter values for the target area and non-target area of the multi-frame image.
  • the preset frame image is segmented, and then the target area and the non-target area can be separately used according to the characteristics of the target area and the non-target area.
  • the noise reduction process not only ensures the overall noise reduction effect of the image and the purity of the target area, but also better retains the detailed information of the non-target area, improves the quality of the captured image, and improves the user experience.
  • FIG. 1 is a schematic flowchart of a night scene image processing method provided by an embodiment of the present application
  • FIG. 2 is a schematic flowchart of another night scene image processing method provided by an embodiment of the present application.
  • FIG. 3 is a schematic structural diagram of a night scene image processing device provided by an embodiment of the present application.
  • FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • FIG. 5 is a schematic structural diagram of another electronic device provided by an embodiment of the present application.
  • the embodiments of the present application aim to solve the problem that when using traditional noise reduction methods to reduce noise in night scene images in the related art, the noise reduction effect of the sky area can not be guaranteed, and the details of other areas are kept good, which affects the user experience.
  • the night scene image processing method provided in the embodiment of the present application can sequentially acquire multiple frames of images according to a preset exposure compensation mode, and use the preset image recognition model to perform recognition processing on the preset frame images in the multiple frame images to Determine the target area and non-target area in the preset frame image, and then use different noise reduction parameter values to perform noise reduction processing on the target area and non-target area of the multi-frame image to generate the target image, where the target area corresponds to The value of the noise reduction parameter is greater than the value of the noise reduction parameter corresponding to the non-target area. Therefore, by using the preset image recognition model, the preset frame image is segmented, and then the target area and the non-target area can be separately used according to the characteristics of the target area and the non-target area.
  • the noise reduction process not only ensures the overall noise reduction effect of the image and the purity of the target area, but also better retains the detailed information of the non-target area, improves the quality of the captured image, and improves the user experience.
  • FIG. 1 is a schematic flowchart of a night scene image processing method provided by an embodiment of the present application.
  • the night scene image processing method includes the following steps:
  • Step 101 sequentially acquire multiple frames of images.
  • the preset exposure compensation mode may include parameters such as exposure value, sensitivity, and exposure duration corresponding to each frame of image.
  • multiple frames of images may be collected in sequence according to the parameters in the preset exposure compensation mode.
  • the overall brightness of the night scene shooting scene is low, you can adjust the exposure time corresponding to each frame of image and collect multiple frames of images with different exposure values to obtain images of different dynamic ranges and synthesize them.
  • the image has a higher dynamic range, improving the overall brightness and quality of the image.
  • the preset exposure compensation mode can be determined in real time according to the current shooting scene to obtain the best shooting effect. That is, in a possible implementation form of the embodiment of the present application, before the foregoing step 101, it may include:
  • the preset exposure compensation mode is determined according to the illuminance of the current shooting scene and the current shaking degree of the camera module.
  • the exposure value corresponding to each frame of image may be different.
  • the exposure value corresponding to each frame of image can be appropriately reduced; when the illuminance of the current shot is small, the exposure value corresponding to each frame of image can be appropriately increased.
  • the current jitter level of the electronic device that is, the current jitter level of the camera module can be determined by acquiring the current gyro-sensor information of the electronic device.
  • Gyroscope is also called angular velocity sensor, which can measure the angular velocity of rotation when the physical quantity is deflected and tilted.
  • the gyroscope can measure the rotation and deflection movement well, so that the actual movement of the user can be accurately analyzed and judged.
  • the gyro information (gyro information) of the electronic device may include the motion information of the electronic device in three dimensions of the three-dimensional space.
  • the three dimensions of the three-dimensional space may be represented as three directions of the X axis, the Y axis, and the Z axis, respectively , X-axis, Y-axis, and Z-axis are perpendicular to each other.
  • the current shaking degree of the camera module may be determined according to the current gyro information of the electronic device.
  • the absolute value threshold of the gyro movement in three directions can be preset, and the camera module is determined according to the relationship between the acquired absolute value of the current absolute value of the gyro movement in the three directions and the preset threshold The current degree of jitter.
  • the preset threshold values are the first threshold A, the second threshold B, and the third threshold C, and A ⁇ B ⁇ C, and the sum of the absolute values of the currently acquired gyro motions in three directions is S .
  • S ⁇ A determine the current jitter level of the camera module as "no jitter”; if A ⁇ S ⁇ B, then determine the current jitter level of the camera module as "slight jitter”; if B ⁇ S ⁇ C, It can be determined that the current jitter level of the camera module is "small jitter”; if S> C, the current jitter level of the camera module can be determined to be "large jitter”.
  • the above examples are only exemplary and cannot be considered as limitations to the present application.
  • the number of thresholds and specific values of the thresholds can be preset according to actual needs, and the mapping relationship between the gyro information and the jitter of the camera module can be preset according to the relationship between the gyro information and the thresholds.
  • the number of collected images and the sensitivity of the collected images will affect the overall shooting duration. If the shooting duration is too long, it may cause the camera module to shake more during handheld shooting, thereby affecting the image quality. Therefore, the number of captured images, the corresponding sensitivity of each frame of image, and the exposure duration of each frame of the image to be captured can be determined according to the current degree of camera module shake, so that the shooting duration is controlled within an appropriate range to avoid shake The degree exacerbates the ghosts introduced.
  • the determining the preset exposure compensation mode according to the illumination intensity of the current shooting scene and the current shaking degree of the camera module includes:
  • the exposure duration of each frame image is determined according to the sensitivity of each frame image and the target exposure value of each frame image.
  • the exposure value refers to the amount of light passing through the lens during the exposure time.
  • the sensitivity also known as the ISO value, refers to an index that measures the sensitivity of the negative film to light. For a film with lower sensitivity, a longer exposure time is required to achieve the same imaging as a film with higher sensitivity.
  • the sensitivity of the digital camera is an indicator similar to the film sensitivity.
  • the ISO of the digital camera can be adjusted by adjusting the sensitivity of the photosensitive device or combining the photosensitive points, that is, by increasing the light sensitivity of the photosensitive device or Merging several adjacent photosensitive points to achieve the purpose of improving ISO. It should be noted that, whether it is digital or film photography, in order to reduce the exposure time, the use of relatively high sensitivity usually introduces more noise, resulting in reduced image quality.
  • the exposure time refers to the time that light passes through the lens.
  • the exposure value is related to the aperture size, exposure duration and sensitivity.
  • the aperture is the clear aperture, which determines the amount of light passing through per unit time.
  • the light metering module in the camera module may be used to obtain the light intensity of the current shooting scene, and the Auto Exposure Control (AEC) algorithm is used to determine the exposure value corresponding to the current light intensity.
  • AEC Auto Exposure Control
  • the exposure value of each frame of images may be different to obtain images with different dynamic ranges, so that the synthesized image has a higher dynamic range, and the overall brightness and quality of the image are improved. That is, when each frame of image is collected, different exposure compensation strategies can be adopted, and the target exposure value corresponding to each frame of image can be determined according to the exposure compensation strategy and the current light intensity.
  • different exposure compensation strategies may be adopted for each frame image by preset exposure compensation strategies, so that each frame image corresponds to a different exposure value, so as to obtain images with different dynamic ranges.
  • the preset exposure compensation strategy refers to a combination of exposure values (Exposure Values, EV for short) preset for each frame of images.
  • exposure value does not refer to an accurate value, but refers to "a combination of all camera apertures and exposure durations that can give the same exposure amount.”
  • Sensitivity, aperture, and exposure duration determine the exposure of the camera. Different parameter combinations can produce equal exposure, that is, the EV values of these different combinations are the same.
  • the sensitivity is the same, use 1/125
  • the combination of the exposure duration of seconds and the aperture of F / 11 is the same as the combination of using the exposure time of 1/250 seconds and the shutter of F / 8.0, that is, the EV value is the same.
  • the EV value is 0, it means the exposure amount obtained when the sensitivity is 100, the aperture factor is F / 1, and the exposure time is 1 second; the exposure amount is increased by one step, that is, the exposure time is doubled, or the sensitivity is doubled , Or the aperture increases by one gear, the EV value increases by 1, that is, the exposure corresponding to 1EV is twice the exposure corresponding to 0EV.
  • Table 1 it is the corresponding relationship with the EV value when the exposure time, aperture, and sensitivity change independently.
  • EV is often used to indicate a level difference on the exposure scale.
  • Many cameras allow exposure compensation to be set, usually expressed in EV.
  • EV refers to the difference between the exposure corresponding to the metering data of the camera and the actual exposure.
  • the exposure compensation of + 1EV refers to an increase in exposure relative to the exposure corresponding to the metering data of the camera, that is, the actual The exposure is twice the exposure corresponding to the camera's metering data.
  • the EV value corresponding to the determined reference exposure amount may be preset to 0, and + 1EV means to increase the exposure by one level, that is, the exposure amount is twice the reference exposure amount.
  • + 2EV refers to an increase of two exposures, that is, the exposure is 4 times the reference exposure
  • -1EV refers to a reduction of exposure, that is, the exposure is 0.5 times the reference exposure, and so on.
  • the EV value range corresponding to the preset exposure compensation strategy may be [+ 1, + 1, + 1, + 1,0, -3, -6].
  • the frame with the exposure compensation strategy of + 1EV can solve the noise problem. Time-domain noise reduction is performed through the frame with higher brightness to suppress the noise while improving the details of the dark part; the frame with the exposure compensation strategy of -6EV can solve the highlight For the problem of exposure, retain the details of the highlight area; the frames with the exposure compensation strategy of 0EV and -3EV can be used to maintain the transition between highlights and dark areas, and maintain a good light and dark transition effect.
  • each EV value corresponding to the preset exposure compensation strategy can be specifically set according to actual needs, or it can be determined according to the set EV value range and the principle that the differences between the EV values are equal Therefore, the embodiment of the present application does not limit this.
  • the corresponding exposure value of each frame of image Target exposure value after the reference exposure value is determined by the ACE algorithm according to the illuminance of the current shooting scene, the corresponding exposure value of each frame of image Target exposure value.
  • the sensitivity of each frame of image refers to the lowest sensitivity determined according to the current degree of shaking of the camera module and adapted to the current degree of shaking.
  • multiple frames of lower sensitivity images can be collected at the same time, and the collected multiple frames of images can be combined to generate a target image, which can not only improve the dynamic range and The overall brightness, and by controlling the value of sensitivity, effectively suppress the noise in the image and improve the quality of the image captured at night.
  • the sensitivity of each frame of image may be determined according to the current shaking degree of the camera module, so that the shooting duration is controlled within an appropriate range. Specifically, if the current jitter of the camera module is small, the sensitivity of each frame of image can be appropriately compressed to a small value to effectively suppress the noise of each frame of image and improve the quality of the captured image; if the camera module If the current degree of jitter is large, the sensitivity of each frame of image can be appropriately increased to a larger value to shorten the shooting time and avoid the introduction of ghosts due to the increased degree of jitter.
  • the sensitivity may be determined to be a smaller value to try to obtain a higher quality Image, for example, the sensitivity is determined to be 100; if it is determined that the camera module's current jitter level is "slight jitter", it may be determined that the current shooting mode may be handheld.
  • the sensitivity is determined to be 200; if the current jitter of the camera module is determined to be "small jitter", you can further reduce the number of images to be acquired and further increase the sensitivity to reduce the shooting duration, such as determining the sensitivity If it is determined that the current jitter of the camera module is "large jitter", it can be determined that the current jitter is too large. At this time, the sensitivity can be further increased to reduce the shooting duration, for example, the sensitivity is determined to be 250.
  • the number of collected images also affects the overall shooting duration.
  • the sensitivity of each frame of image can be appropriately compressed to a smaller value to effectively suppress the noise of each frame of image and improve shooting The quality of the image; if the camera module currently has a large degree of jitter, images of fewer frames can be collected, and the sensitivity of each frame of the image can be appropriately increased to a larger value to shorten the shooting time.
  • the current shooting mode may be a tripod shooting mode. At this time, more frames of images may be collected and the sensitivity is determined to be a smaller value .
  • the number of multi-frame images is 17 frames, and the sensitivity is 100; if the current jitter level of the camera module is determined to be "slight jitter", it can be determined that the current shooting mode may be At this time, you can collect images with fewer frames and determine the sensitivity to a larger value to reduce the shooting time.
  • the number of multi-frame images is 7 frames and the sensitivity is 200; if the current camera module is determined If the jitter level is "small jitter", you can further reduce the number of multi-frame images, and further increase the sensitivity to reduce the shooting time. For example, if the number of multi-frame images is 5 frames, the sensitivity is 220; if the camera mode is determined If the current jitter level of the group is "large jitter", you can determine that the current jitter level is too large. At this time, you can further reduce the number of multi-frame images and further increase the sensitivity to reduce the shooting duration, such as determining the multi-frame image The number is 3 frames and the sensitivity is 250.
  • multiple sets of exposure compensation strategies can also be preset to determine the preset exposure that matches the number of multi-frame images according to the number of multi-frame images Compensation strategy.
  • both the number and sensitivity of the images to be acquired can be changed at the same time, or one of them can be changed to obtain the optimal solution.
  • the mapping relationship between the degree of shaking of the camera module and the number of images to be collected and the sensitivity corresponding to the images to be collected per frame can be preset according to actual needs.
  • each frame after determining the target exposure value of each frame image and the sensitivity of each frame image in the multi-frame image, each frame can be determined according to the sensitivity of each frame image and the target exposure value of each frame image The exposure time of the image.
  • Step 102 Using a preset image recognition model, perform recognition processing on the preset frame images in the multi-frame images to determine target areas and non-target areas in the preset frame images.
  • the preset image recognition model refers to an image segmentation model obtained by training a large amount of annotated night scene image data.
  • the target area refers to a flat area with abnormal brightness that requires high intensity noise reduction.
  • the non-target area refers to an area other than the target area in the preset frame image, which is generally a texture area with moderate brightness and containing more detailed information. That is, in a possible implementation form of the embodiment of the present application, the above step 102 may include:
  • a preset image recognition model is used to perform recognition processing on the preset frame images in the multi-frame image to determine that the area with abnormal brightness in the preset frame image is the target area.
  • the area with abnormal brightness may refer to an area where the difference in brightness from the area around it exceeds a preset threshold.
  • the brightness of the sky area is usually very dark compared to other areas, so the sky area in the night scene image can be determined as the area with abnormal brightness, that is, the target area; compared with other areas in the image
  • the brightness of the reflective area is usually very large, so the reflective area in the image can be determined as the area with abnormal brightness, that is, the target area; for another example, the halo area near the light source is usually lower in brightness than the light source area, so The halo area in the image can also be determined as the area with abnormal brightness, that is, the target area.
  • the preset threshold value of the brightness difference may be preset according to actual needs, which is not limited in the embodiment of the present application.
  • the target area and the non-target area can be determined from the preset frame image through the preset image recognition model, and the target area and the non-target area are respectively reduced by using different noise reduction parameters noise.
  • the preset image recognition model can be generated offline and integrated into the electronic device.
  • the trained image recognition model can analyze the image data of the input image recognition model, and determine the label information of each pixel according to the pixel value of each pixel and the relationship between the pixel value of each pixel, and The pixel points of the input image are labeled to obtain a mask that can distinguish the target area from the non-target area, thereby determining the target area and the non-target area.
  • the preset rule when labeling a large amount of collected night scene image data, the preset rule may be “label the pixel corresponding to the target area as 1, and the pixel corresponding to the non-target area as 0.”
  • the image recognition model obtained by training the labeled image data can analyze the image data of the input image recognition model, and according to the relationship between the pixel value of each pixel and the pixel value of each pixel, the same
  • the rule marks the pixels of the input image, that is, the pixels whose label information is "1" are the target area, and the pixels whose label information is "0" are the non-target area.
  • the preset frame image in the multi-frame image is input into the preset image recognition model, and the preset image recognition model can perform recognition processing on the preset frame image, that is, according to each pixel in the preset frame image
  • the relationship between the pixel value of each pixel and the pixel value of each pixel determines the label information of each pixel, and then determines the target area and non-target area in the preset frame image.
  • one frame image can be selected from the multiple frame images for recognition processing through a preset image recognition model, and the multiple frame images can be obtained according to the recognition result.
  • Target area and non-target area The selected frame image is the preset frame image. That is, in a possible implementation form of the embodiment of the present application, before the above step 102, it may further include:
  • the preset frame image is determined according to the exposure value corresponding to each frame image in the multiple frame images.
  • an image with high brightness and sharpness may be used as a preset frame image to obtain an optimal segmentation effect.
  • the larger the exposure value the higher the brightness and sharpness of the image, so the corresponding image with the largest exposure value can be preset as a preset frame image, that is, the exposure value can be determined according to the exposure value corresponding to each frame image
  • the largest image is the preset frame image. If the corresponding image with the largest exposure value has multiple frames, one frame can be randomly selected from the multiple frames with the largest exposure value as the preset frame image.
  • Step 103 Use different noise reduction parameter values to perform noise reduction processing on the target area and the non-target area of the multi-frame image to generate a target image, wherein the value of the noise reduction parameter corresponding to the target area is greater than the The noise reduction parameter value corresponding to the non-target area.
  • different noise reduction parameter values can be used according to the characteristics of the target area and the non-target area. Perform noise reduction on the target area and non-target area separately.
  • the target area contains more noise and less detailed information
  • the non-target area contains less noise and rich detailed information.
  • a larger noise reduction parameter value can be used for the target area, such as increasing the noise reduction intensity to remove a large amount of noise in the target area to achieve the best noise reduction effect; and
  • a smaller noise reduction parameter value can be used, such as reducing the noise reduction intensity, so as to reduce the noise while reducing the damage of the image detail information by the noise reduction processing and improving the image quality.
  • the noise reduction processing of the multi-frame images may be performed in a mixed manner of spatial domain noise reduction and temporal domain noise reduction.
  • the spatial domain noise reduction uses the spatial correlation within a single frame image for noise reduction
  • the temporal domain noise reduction uses the temporal correlation of multiple frame images for noise reduction.
  • the most common method of spatial domain filtering is low-pass filtering.
  • the low-pass filtering method achieves the purpose of noise reduction by filtering out the high-frequency part of the image signal, but because the edge and jumping part of the image are also in the high-frequency region, it is easy to cause the image edge and The jump part is blurred, causing damage to the image detail information concentrated in the high frequency of the signal.
  • the spatial domain noise reduction does not consider the time domain information, and the noise at the same position between the frames is random, it is easy to cause the image content between adjacent frames to change after the noise reduction.
  • time-domain noise reduction is the multi-image averaging method. Because the image content has a strong correlation between frames, the image content changes continuously within each frame, and noise always appears randomly between frames. With correlation, the noise is discontinuous in each frame, so time-domain noise reduction takes advantage of the characteristics of noise between frames, which can effectively remove noise and protect the details of the image.
  • time-domain noise reduction takes advantage of the characteristics of noise between frames, which can effectively remove noise and protect the details of the image.
  • the simple time-domain noise reduction is prone to matching failures or errors, and the phenomenon of residual noise or "ghosting" occurs. Therefore, the noise reduction process of multi-frame images can be effectively suppressed by a combination of spatial noise reduction and temporal noise reduction, and the details such as image edges and textures can be better preserved.
  • the night scene image processing method provided in the embodiment of the present application can sequentially acquire multiple frames of images according to a preset exposure compensation mode, and use the preset image recognition model to perform recognition processing on the preset frame images in the multiple frame images to Determine the target area and non-target area in the preset frame image, and then use different noise reduction parameter values to perform noise reduction processing on the target area and non-target area of the multi-frame image to generate the target image, where the target area corresponds to The value of the noise reduction parameter is greater than the value of the noise reduction parameter corresponding to the non-target area. Therefore, by using the preset image recognition model, the preset frame image is segmented, and then the target area and the non-target area can be separately used according to the characteristics of the target area and the non-target area.
  • the noise reduction process not only ensures the overall noise reduction effect of the image and the purity of the target area, but also better retains the detailed information of the non-target area, improves the quality of the captured image, and improves the user experience.
  • the target area can also be feathered to determine the transition area between the target area and the non-target area, and different noise reduction parameter values are used to target the target area, non-target area,
  • the transition area is subjected to noise reduction processing to achieve a natural transition of the target noise reduction effect and further improve the image quality.
  • FIG. 2 is a schematic flowchart of another night scene image processing method provided by an embodiment of the present application.
  • the night scene image processing method includes the following steps:
  • Step 201 sequentially acquire multiple frames of images.
  • Step 202 Using a preset image recognition model, perform a recognition process on the preset frame images in the multi-frame images to determine target areas and non-target areas in the preset frame images.
  • Step 203 Perform feathering on the target area to determine a transition area between the target area and the non-target area.
  • the feathering process refers to blurring the edges of the image so that the edges of the image can achieve a hazy effect.
  • Feathering the target area means blurring the edges of the target area to achieve the effect of a natural transition between the target area and the non-target area.
  • the hazy range of the edge that is, the size of the transition area can be controlled by adjusting the feather radius value.
  • feathering the target area is a process of re-determining the pixel values of pixels near the edge of the target area within a feathering radius by a certain method.
  • feathering through mean smoothing refers to re-determining the mean value of a pixel's neighborhood pixels as the pixel value of that pixel. For example, if the neighborhood size is 11 ⁇ 11, the pixel value of pixel A is 100, and the average pixel value of the pixels in the 11 ⁇ 11 neighborhood is 85, then the pixel value of pixel A after feathering is 85.
  • Step 204 Using different noise reduction parameter values, respectively perform noise reduction processing on the target area, the transition area and the non-target area of the multi-frame image to generate a target image.
  • different noise reduction parameter values can be used to respectively reduce the target area, the transition area and the non-target area in the multi-frame image Processing, the target image has been generated.
  • the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the transition area, and the noise reduction parameter value corresponding to the transition area is greater than the noise reduction parameter value corresponding to the non-target area.
  • the noise reduction parameter value corresponding to each area can be determined according to the noise intensity contained in the image to obtain the best noise reduction effect.
  • the noise level in the collected multi-frame images is related to the current shooting scene.
  • the sensitivity when shooting, the illumination of the shooting scene, and the details in the scene can all affect the noise level in the collected image.
  • the reference noise reduction parameters can be determined according to the current shooting scene, and then
  • the noise reduction parameter value corresponding to each area is determined according to the reference noise reduction parameter. That is, in a possible implementation form of the embodiment of the present application, before the above step 204, it may further include:
  • the preset weight value corresponding to each region and the reference noise reduction parameter determine the noise reduction parameter values currently corresponding to the target region, the transition region, and the non-target region, respectively.
  • the current shooting scene is closely related to the noise level in the collected multi-frame images. Therefore, the noise level in the multi-frame images can be estimated according to the current shooting scene, and the noise level can be determined in accordance with the estimated noise level. Adapted reference noise reduction parameters.
  • the reference noise reduction parameter For example, if the current sensitivity is small, or the details of the scene to be captured are rich, you can determine that the noise level in the acquired multi-frame image is low, and you can determine the reference noise reduction parameter to be small Value; if the current sensitivity is large, or there is little detail information in the scene to be shot, you can determine that the noise level in the acquired multi-frame image is high, you can determine the reference noise reduction parameter as a larger value.
  • a weight value during noise reduction may be preset for each area, and the noise reduction parameter value corresponding to each area is determined according to the weight value and the reference noise reduction parameter.
  • the reference noise reduction parameter can be used to reduce noise in the non-target area, that is, the weight value corresponding to the non-target area can be 1; and the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the non-target area, namely The weight value corresponding to the target area needs to be greater than 1, such as 1.5; the corresponding noise reduction parameter value of the transition area needs to be between the noise reduction parameter value corresponding to the target area and the noise reduction parameter value corresponding to the non-target area, such as It is 1.2.
  • the above examples are only exemplary and cannot be considered as limitations to the present application.
  • the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the transition area
  • the noise reduction parameter value corresponding to the transition area is greater than the noise reduction parameter value corresponding to the non-target area
  • the preset corresponding to each area To obtain the best noise reduction effect.
  • the night scene image processing method provided in the embodiment of the present application can sequentially acquire multiple frames of images according to a preset exposure compensation mode, and use the preset image recognition model to perform recognition processing on the preset frame images in the multiple frame images to Determine the target area and non-target area in the preset frame image, and then feather the target area to determine the transition area between the target area and the non-target area and then use different noise reduction parameter values to target the multi-frame image
  • the area, the transition area and the non-target area are respectively subjected to noise reduction processing to generate a target image, wherein the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the transition area, and the noise reduction parameter value corresponding to the transition area is greater than the non-target The noise reduction parameter value corresponding to the area.
  • the preset frame image is segmented, and the identified target area is feathered to determine the transition area, and then different noise reduction parameter values can be used to target the target
  • the area, transition area and non-target area are subjected to noise reduction processing, so as to not only ensure the overall noise reduction effect of the image and the purity of the target area, but also retain the detailed information of the non-target area, and realize the target area and non-target area
  • the natural transition of the noise reduction effect between regions further improves the quality of the captured image and improves the user experience.
  • the present application also proposes a night scene image processing device.
  • FIG. 3 is a schematic structural diagram of a night view image processing device provided by an embodiment of the present application.
  • the night view image processing device 30 includes:
  • the acquisition module 31 is used to sequentially acquire multiple frames of images according to a preset exposure compensation mode
  • the first determining module 32 is configured to use a preset image recognition model to perform recognition processing on the preset frame images in the multi-frame images to determine target areas and non-target areas in the preset frame images;
  • the noise reduction module 33 is configured to use different noise reduction parameter values to respectively perform noise reduction processing on the target area and the non-target area of the multi-frame image to generate a target image, wherein the noise reduction parameter corresponding to the target area The value is greater than the value of the noise reduction parameter corresponding to the non-target area.
  • the night scene image processing apparatus provided by the embodiments of the present application may be configured in any electronic device to perform the aforementioned night scene image processing method.
  • the night scene image processing device can sequentially acquire multiple frames of images according to a preset exposure compensation mode, and use the preset image recognition model to perform recognition processing on the preset frame images in the multiple frame images to Determine the target area and non-target area in the preset frame image, and then use different noise reduction parameter values to perform noise reduction processing on the target area and non-target area of the multi-frame image to generate the target image, where the target area corresponds to The value of the noise reduction parameter is greater than the value of the noise reduction parameter corresponding to the non-target area. Therefore, by using the preset image recognition model, the preset frame image is segmented, and then the target area and the non-target area can be separately used according to the characteristics of the target area and the non-target area.
  • the noise reduction process not only ensures the overall noise reduction effect of the image and the purity of the target area, but also better retains the detailed information of the non-target area, improves the quality of the captured image, and improves the user experience.
  • the above night scene image processing device 30 further includes:
  • the second determination module is used to determine the preset exposure compensation mode according to the illumination of the current shooting scene and the current shaking degree of the camera module.
  • the above-mentioned second determination module is specifically used for:
  • the exposure duration of each frame image is determined according to the sensitivity of each frame image and the target exposure value of each frame image.
  • the above-mentioned second determination module is also used to:
  • the target exposure value of each image in the multi-frame image is determined.
  • the foregoing night scene image processing device 30 further includes:
  • the third determining module is configured to determine the preset frame image according to the exposure value corresponding to each frame image in the multiple frame images.
  • the foregoing night scene image processing device 30 further includes:
  • the fourth determination module is used to feather the target area to determine the transition area between the target area and the non-target area.
  • the above noise reduction module 33 is specifically used for:
  • the target area, the transition area and the non-target area of the multi-frame image are respectively subjected to noise reduction processing.
  • the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the transition area, and the noise reduction parameter value corresponding to the transition area is greater than the corresponding non-target area. Noise reduction parameter value.
  • the foregoing night scene image processing device 30 further includes:
  • the fifth determination module is used to determine the reference noise reduction parameter according to the current shooting scene
  • the sixth determining module is configured to determine the noise reduction parameter values corresponding to the target area and the non-target area according to the preset weight value corresponding to each area and the reference noise reduction parameter, respectively.
  • the above-mentioned first determination module 32 is specifically used for:
  • a preset image recognition model is used to perform recognition processing on the preset frame images in the multi-frame image to determine that the area with abnormal brightness in the preset frame image is the target area.
  • the night scene image processing device can sequentially acquire multiple frames of images according to a preset exposure compensation mode, and use the preset image recognition model to perform recognition processing on the preset frame images in the multiple frame images to Determine the target area and non-target area in the preset frame image, and then feather the target area to determine the transition area between the target area and the non-target area and then use different noise reduction parameter values to target the multi-frame image
  • the area, the transition area and the non-target area are respectively subjected to noise reduction processing to generate a target image, wherein the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the transition area, and the noise reduction parameter value corresponding to the transition area is greater than the non-target The noise reduction parameter value corresponding to the area.
  • the preset frame image is segmented, and the identified target area is feathered to determine the transition area, and then different noise reduction parameter values can be used to target the target
  • the area, transition area and non-target area are subjected to noise reduction processing, so as to not only ensure the overall noise reduction effect of the image and the purity of the target area, but also retain the detailed information of the non-target area, and realize the target area and non-target area
  • the natural transition of the noise reduction effect between regions further improves the quality of the captured image and improves the user experience.
  • the present application also proposes an electronic device.
  • FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • the electronic device 200 includes a camera module 201, a memory 210, a processor 220, and a computer program stored on the memory and executable on the processor.
  • the processor executes the program, the The night scene image processing method described in the embodiment of the present application.
  • the electronic device 200 provided by the embodiment of the present application may further include:
  • the memory 210 and the processor 220 are connected to the bus 230 of different components (including the memory 210 and the processor 220).
  • the memory 210 stores a computer program.
  • the processor 220 executes the program, the night scene image processing described in the embodiments of the present application is implemented method.
  • the bus 230 represents one or more of several types of bus structures, including a memory bus or a memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of a variety of bus structures.
  • these architectures include, but are not limited to, industry standard architecture (ISA) bus, micro channel architecture (MAC) bus, enhanced ISA bus, video electronics standards association (VESA) local bus, and peripheral component interconnection ( PCI) bus.
  • ISA industry standard architecture
  • MAC micro channel architecture
  • VESA video electronics standards association
  • PCI peripheral component interconnection
  • the electronic device 200 typically includes a variety of electronic device readable media. These media may be any available media that can be accessed by the electronic device 200, including volatile and nonvolatile media, removable and non-removable media.
  • the memory 210 may also include computer system readable media in the form of volatile memory, such as random access memory (RAM) 240 and / or cache memory 250.
  • the electronic device 200 may further include other removable / non-removable, volatile / nonvolatile computer system storage media.
  • the storage system 260 may be used to read and write non-removable, non-volatile magnetic media (not shown in FIG. 5 and is generally referred to as a "hard disk drive").
  • each drive may be connected to the bus 230 through one or more data media interfaces.
  • the memory 210 may include at least one program product, and the program product has a set of (eg, at least one) program modules configured to perform the functions of the embodiments of the present application.
  • a program / utility tool 280 having a set of (at least one) program modules 270 may be stored in, for example, the memory 210.
  • Such program modules 270 include—but are not limited to—operating system, one or more application programs, and other programs Modules and program data, each of these examples or some combination may include the implementation of the network environment.
  • the program module 270 generally performs the functions and / or methods in the embodiments described in this application.
  • the electronic device 200 may also communicate with one or more external devices 290 (eg, keyboard, pointing device, display 291, etc.), and may also communicate with one or more devices that enable a user to interact with the electronic device 200, and / or This enables the electronic device 200 to communicate with any device (such as a network card, modem, etc.) that communicates with one or more other computing devices. Such communication may be performed through an input / output (I / O) interface 292.
  • the electronic device 200 can also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 293.
  • networks such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet
  • the network adapter 293 communicates with other modules of the electronic device 200 through the bus 230. It should be understood that although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 200, including but not limited to: microcode, device driver, redundant processing unit, external disk drive array, RAID system, tape drive And data backup storage system.
  • the processor 220 runs programs stored in the memory 210 to execute various functional applications and data processing.
  • the electronic device provided in the embodiment of the present application can execute the night scene image processing method as described above, collect multiple frames of images in sequence according to the preset exposure compensation mode, and use the preset image recognition model to Recognize the preset frame image to determine the target area and non-target area in the preset frame image, and then use different noise reduction parameter values to perform noise reduction on the target area and non-target area of the multi-frame image, respectively, to Generate a target image, where the noise reduction parameter value corresponding to the target area is greater than the noise reduction parameter value corresponding to the non-target area. Therefore, by using the preset image recognition model, the preset frame image is segmented, and then the target area and the non-target area can be separately used according to the characteristics of the target area and the non-target area.
  • the noise reduction process not only ensures the overall noise reduction effect of the image and the purity of the target area, but also better retains the detailed information of the non-target area, improves the quality of the captured image, and improves the user experience.
  • the present application also proposes a computer-readable storage medium.
  • the computer-readable storage medium has stored thereon a computer program, and when the program is executed by the processor, to implement the night scene image processing method described in the embodiments of the present application.
  • another embodiment of the present application provides a computer program, which when executed by a processor, implements the night scene image processing method described in the embodiments of the present application.
  • the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium.
  • the computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination of the above. More specific examples of computer-readable storage media (non-exhaustive list) include: electrical connections with one or more wires, portable computer diskettes, hard drives, random access memory (RAM), read-only memory (ROM), Erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.
  • the computer-readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in combination with an instruction execution system, apparatus, or device.
  • the computer-readable signal medium may include a data signal that is propagated in baseband or as part of a carrier wave, in which computer-readable program code is carried. This propagated data signal can take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above.
  • the computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, and the computer-readable medium may send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. .
  • the program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
  • the computer program code for performing the operations of the present invention can be written in one or more programming languages or a combination thereof.
  • the programming languages include object-oriented programming languages such as Java, Smalltalk, C ++, and also include conventional Procedural programming language-such as "C" language or similar programming language.
  • the program code may be completely executed on the user electronic device, partly executed on the user electronic device, executed as an independent software package, partly executed on the user electronic device and partly executed on the remote electronic device, or completely executed on the remote electronic device or Executed on the server.
  • the remote electronic devices may be connected to user electronic devices through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to external electronic devices (eg, using Internet services Provider to connect via the Internet).
  • LAN local area network
  • WAN wide area network
  • Internet services Provider to connect via the Internet

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Studio Devices (AREA)

Abstract

一种夜景图像处理方法、装置、电子设备及存储介质，属于成像技术领域。其中，方法包括：根据预设的曝光补偿模式，依次采集多帧图像；利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域；采用不同的降噪参数值，对多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过这种夜景图像处理方法，不仅保证了图像整体的降噪效果和目标区域的纯净度，而且较好的保留了非目标区域的细节信息，提高了拍摄图像的质量，改善了用户体验。

Description

夜景图像处理方法、装置、电子设备及存储介质
相关申请的交叉引用
本申请要求OPPO广东移动通信有限公司于2018年11月22日提交的、发明名称为“夜景图像处理方法、装置、电子设备及存储介质”的、中国专利申请号“201811399541.0”的优先权。
技术领域
本申请涉及成像技术领域，尤其涉及一种夜景图像处理方法、装置、电子设备及存储介质。
背景技术
手机和平板电脑都内置有摄像头，并且随着移动终端处理能力的增强以及摄像头技术的发展，内置摄像头的性能越来越强大，拍摄图像的质量也越来越高。现在移动终端的操作简单又便于携带，在日常生活中人们使用智能手机和平板电脑等移动终端拍照已经成为一种常态。
智能移动终端在给人们的日常拍照带来便捷的同时，人们对拍摄的图像质量的要求也越来越高。目前，为了满足人们对图像质量的要求，在夜景拍摄场景中，移动终端通常通过采集多帧曝光时长不同的图像并进行融合的方式，进行夜景图像的拍摄。但是，在这种拍摄方式中，如何对夜景天空区域进行降噪是一大难点。
发明内容
本申请提出的夜景图像处理方法、装置、电子设备及存储介质，用于解决相关技术中采用传统降噪方法对夜景拍摄图像进行降噪时，无法既保证天空区域的降噪效果，又保证其他区域的细节信息保持良好，影响用户体验的问题。
本申请一方面实施例提出的夜景图像处理方法，包括：根据预设的曝光补偿模式，依次采集多帧图像；利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域；采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，所述目标区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
本申请另一方面实施例提出的夜景图像处理装置，包括：采集模块，用于根据预设的 曝光补偿模式，依次采集多帧图像；确定模块，用于利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域；降噪模块，用于采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，所述目标区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
本申请再一方面实施例提出的电子设备，其包括：所述摄像模组、存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器执行所述程序时实现如前所述的夜景图像处理方法。
本申请再一方面实施例提出的计算机可读存储介质，其上存储有计算机程序，其特征在于，所述程序被处理器执行时实现如前所述的夜景图像处理方法。
本申请又一方面实施例提出的计算机程序，该程序被处理器执行时，以实现本申请实施例所述的夜景图像处理方法。
本申请实施例提供的夜景图像处理方法、装置、电子设备、计算机可读存储介质及计算机程序，可以根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，进而采用不同的降噪参数值，对多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，之后即可根据目标区域与非目标区域的特性，采用不同的降噪参数值分别对目标区域与非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，而且较好的保留了非目标区域的细节信息，提高了拍摄图像的质量，改善了用户体验。
本申请附加的方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本申请的实践了解到。
附图说明
本申请上述的和/或附加的方面和优点从下面结合附图对实施例的描述中将变得明显和容易理解，其中：
图1为本申请实施例所提供的一种夜景图像处理方法的流程示意图；
图2为本申请实施例所提供的另一种夜景图像处理方法的流程示意图；
图3为本申请实施例所提供的一种夜景图像处理装置的结构示意图；
图4为本申请实施例所提供的一种电子设备的结构示意图；
图5为本申请实施例所提供的另一种电子设备的结构示意图。
具体实施方式
下面详细描述本申请的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的要素。下面通过参考附图描述的实施例是示例性的，旨在用于解释本申请，而不能理解为对本申请的限制。
本申请实施例针对相关技术中采用传统降噪方法对夜景拍摄图像进行降噪时，无法既保证天空区域的降噪效果，又保证其他区域的细节信息保持良好，影响用户体验的问题，提出一种夜景图像处理方法。
本申请实施例提供的夜景图像处理方法，可以根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，进而采用不同的降噪参数值，对多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，之后即可根据目标区域与非目标区域的特性，采用不同的降噪参数值分别对目标区域与非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，而且较好的保留了非目标区域的细节信息，提高了拍摄图像的质量，改善了用户体验。
下面参考附图对本申请提供的夜景图像处理方法、装置、电子设备、存储介质及计算机程序进行详细描述。
图1为本申请实施例所提供的一种夜景图像处理方法的流程示意图。
如图1所示，该夜景图像处理方法，包括以下步骤：
步骤101，根据预设的曝光补偿模式，依次采集多帧图像。
在对夜景图像中的天空区域进行降噪时，由于夜景天空区域的亮度都比较暗，因此在最终图像融合时通常采用曝光时长较长的图像数据，对天空区域进行融合，而曝光时长较长的图像中的噪声通常较高，从而导致夜景天空区域的噪声较高。申请人发现，在采用传统降噪方法对夜景拍摄图像进行降噪时，无法既保证天空区域的降噪效果，又保证其他区域的细节信息保持良好，影响了用户体验。
其中，预设的曝光补偿模式中，可以包括每帧图像对应的曝光值、感光度、曝光时长等参数。在本申请实施例中，可以根据预设的曝光补偿模式中的参数，依次采集多帧图像。
需要说明的是，由于夜景拍摄场景的整体亮度较低，因此可以通过调整每帧图像对应的曝光时长，采集多帧曝光值不同的图像，以获得不同动态范围的图像并进行合成，使得合成后的图像具有更高的动态范围，提高图像的整体亮度和质量。
进一步的，可以根据当前的拍摄场景实时确定预设的曝光补偿模式，以获得最佳的拍摄效果。即在本申请实施例一种可能的实现形式中，上述步骤101之前，可以包括：
根据当前拍摄场景的光照度和摄像模组当前的抖动程度，确定所述预设的曝光补偿模式。
可以理解的是，当前拍摄场景的光照度不同时，每帧图像对应的曝光值可以是不同的。比如，当前拍摄场景的光照度较大时，可以适当降低每帧图像对应的曝光值；当前拍摄的光照度较小时，可以适当提高每帧图像对应的曝光值。
在本申请实施例中，可以通过获取电子设备当前的陀螺仪(Gyro-sensor)信息，确定电子设备当前的抖动程度，即摄像模组当前的抖动程度。
陀螺仪又叫角速度传感器，可以测量物理量偏转、倾斜时的转动角速度。在电子设备中，陀螺仪可以很好的测量转动、偏转的动作，从而可以精确分析判断出使用者的实际动作。电子设备的陀螺仪信息(gyro信息)可以包括电子设备在三维空间中三个维度方向上的运动信息，三维空间的三个维度可以分别表示为X轴、Y轴、Z轴三个方向，其中，X轴、Y轴、Z轴为两两垂直关系。
需要说明的是，在本申请实施例一种可能的实现形式中，可以根据电子设备当前的gyro信息，确定摄像模组当前的抖动程度。电子设备在三个方向上的gyro运动的绝对值越大，则摄像模组的抖动程度越大。具体的，可以预设在三个方向上gyro运动的绝对值阈值，并根据获取到的当前在三个方向上的gyro运动的绝对值之和，与预设的阈值的关系，确定摄像模组的当前的抖动程度。
举例来说，假设预设的阈值为第一阈值A、第二阈值B、第三阈值C，且A<B<C，当前获取到的在三个方向上gyro运动的绝对值之和为S。若S<A，则确定摄像模组当前的抖动程度为“无抖动”；若A<S<B，则可以确定摄像模组当前的抖动程度为“轻微抖动”；若B<S<C，则可以确定摄像模组当前的抖动程度为“小抖动”；若S>C，则可以确定摄像模组当前的抖动程度为“大抖动”。
需要说明的是，上述举例仅为示例性的，不能视为对本申请的限制。实际使用时，可以根据实际需要预设阈值的数量和各阈值的具体数值，以及根据gyro信息与各阈值的关系，预设gyro信息与摄像模组抖动程度的映射关系。
可以理解的是，采集的图像的数量以及采集图像的感光度会影响到整体的拍摄时长，拍摄时长过长，可能会导致手持拍摄时摄像模组的抖动程度加剧，从而影响图像质量。因此，可以根据摄像模组当前的抖动程度，确定采集的图像数量、每帧图像对应的感光度，以及每帧待采集图像对应的曝光时长，以使得拍摄时长控制在合适的范围内，避免抖动程度加剧引入的鬼影。
具体的，在本申请实施例一种可能的实现形式中，所述根据当前拍摄场景的光照度和摄像模组当前的抖动程度，确定所述预设的曝光补偿模式，包括：
根据当前拍摄场景的光照度，确定所述多帧图像中每帧图像的目标曝光值；
根据所述摄像模组当前的抖动程度，确定所述每帧图像的感光度；
根据所述每帧图像的感光度及每帧图像的目标曝光值，确定所述每帧图像的曝光时长。
其中，曝光值，是指曝光时间内通过镜头的光线的数量。
其中，感光度，又称为ISO值，是指衡量底片对于光的灵敏程度的指标。对于感光度较低的底片，需要曝光更长的时间以达到跟感光度较高的底片相同的成像。数码相机的感光度是一种类似于胶卷感光度的一种指标，数码相机的ISO可以通过调整感光器件的灵敏度或者合并感光点来调整，也就是说，可以通过提升感光器件的光线敏感度或者合并几个相邻的感光点来达到提升ISO的目的。需要说明的是，无论是数码或是底片摄影，为了减少曝光时间，使用相对较高的感光度通常会引入较多的噪声，从而导致图像质量降低。
其中，曝光时长，是指光线通过镜头的时间。
需要说明的是，曝光值与光圈大小、曝光时长和感光度有关。其中，光圈也就是通光口径，决定单位时间内光线通过的数量。当每帧图像对应的感光度相同，并且光圈大小相同时，当前拍摄场景的光照度对应的曝光量越大，每帧图像对应的曝光时长越长。
在本申请实施例中，可以利用摄像模组中的测光模块，获取当前拍摄场景的光照度，并利用自动曝光控制(Auto Exposure Control，简称AEC)算法，确定当前光照度对应的曝光值。在采集多帧图像的拍摄模式中，每帧图像的曝光值可以是不同的，以获得具有不同动态范围的图像，使得合成后的图像具有更高的动态范围，提高图像的整体亮度和质量。即可以在采集每帧图像时，采用不同的曝光补偿策略，并根据曝光补偿策略以及当前的光照度，确定出每帧图像对应的目标曝光值。
在本申请实施例中，可以通过预设曝光补偿策略，对每帧图像分别采取不同的曝光补偿策略，使得每帧图像对应于不同的曝光值，以获得具有不同动态范围的图像。
在本申请实施例中，预设的曝光补偿策略，是指为每帧图像分别预设的曝光值(Exposure Value,简称EV)的组合。在曝光值最初的定义中，曝光值并不是指一个准确的数值，而是指“能够给出相同的曝光量的所有相机光圈与曝光时长的组合”。感光度、光圈和曝光时长确定了相机的曝光量，不同的参数组合可以产生相等的曝光量，即这些不同组合的EV值是一样的，比如，在感光度相同的情况下，使用1/125秒曝光时长和F/11的光圈组合，与使用1/250秒曝光时间与F/8.0快门的组合，获得的曝光量是相同的，即EV值是相同的。EV值为0时，是指感光度为100、光圈系数为F/1、曝光时长为1秒时获得的曝光量；曝光量增加一档，即曝光时长增加一倍，或者感光度增加一倍，或者光圈增加 一档，EV值增加1，也就是说，1EV对应的曝光量是0EV对应的曝光量的两倍。如表1所示，为曝光时长、光圈、感光度分别单独变化时，与EV值的对应关系。
表1
Figure PCTCN2019101430-appb-000001
摄影技术进入到数码时代之后，相机内部的测光功能已经非常强大，EV则经常用来表示曝光刻度上的一个级差，许多相机都允许设置曝光补偿，并通常用EV来表示。在这种情况下，EV是指相机测光数据对应的曝光量与实际曝光量的差值，比如+1EV的曝光补偿是指相对于相机测光数据对应的曝光量增加一档曝光，即实际曝光量为相机测光数据对应的曝光量的两倍。
在本申请实施例中，预设曝光补偿策略时，可以将确定的基准曝光量对应的EV值预设为0，+1EV是指增加一档曝光，即曝光量为基准曝光量的2倍，+2EV是指增加两档曝光，即曝光量为基准曝光量的4倍，-1EV是指减少一档曝光，即曝光量为基准曝光量的0.5倍等等。
举例来说，若多帧图像的数量为7帧，则预设的曝光补偿策略对应的EV值范围可以是[+1,+1,+1,+1,0,-3,-6]。其中，曝光补偿策略为+1EV的帧，可以解决噪声问题，通过亮度比较高的帧进行时域降噪，在提升暗部细节的同时抑制噪声；曝光补偿策略为-6EV的帧，可以解决高光过曝的问题，保留高光区域的细节；曝光补偿策略为0EV和-3EV的帧，则可以用于保持高光到暗区之间的过渡，保持较好的明暗过渡的效果。
需要说明的是，预设的曝光补偿策略对应的各EV值既可以是根据实际需要具体设置的，也可以是根据设置的EV值范围，并依据各EV值之间的差值相等的原则求得的，本申请实施例对此不做限定。
在本申请实施例一种可能的实现形式中，根据当前拍摄场景的光照度，通过ACE算法确定出基准曝光值之后，即可根据基准曝光值及预设的曝光补偿策略，确定每帧图像对应的目标曝光值。
在本申请实施例中，每帧图像的感光度，是指根据摄像模组当前的抖动程度，确定的与当前的抖动程度相适应的最低感光度。
需要说明的是，在本申请实施例中，可以通过同时采集多帧感光度较低的图像，并将采集的多帧图像合成以生成目标图像的方式，不仅可以提升夜景拍摄图像的动态范围和整体亮度，并且通过控制感光度的值，有效抑制图像中的噪声，提高夜景拍摄图像的质量。
在本申请实施例中，可以根据摄像模组当前的抖动程度，确定每帧图像的感光度，以使得拍摄时长控制在合适的范围内。具体的，若摄像模组当前的抖动程度较小，则可以将每帧图像的感光度适当压缩为较小的值，以有效抑制每帧图像的噪声、提高拍摄图像的质量；若摄像模组当前的抖动程度较大，则可以将每帧图像的感光度适当提高为较大的值，以缩短拍摄时长，避免抖动程度加剧引入的鬼影。
举例来说，若确定摄像模组当前的抖动程度为“无抖动”，则可以确定当前可能为脚架拍摄模式，此时可以将感光度确定为较小的值，以尽量获得更高质量的图像，比如确定感光度为100；若确定摄像模组当前的抖动程度为“轻微抖动”，则可以确定当前可能为手持拍摄模式，此时可以将感光度确定为较大的值，以降低拍摄时长，比如确定感光度为200；若确定摄像模组当前的抖动程度为“小抖动”，则可以进一步减少待采集图像的数量，并进一步增大感光度，以降低拍摄时长，比如确定感光度为220；若确定摄像模组当前的抖动程度为“大抖动”，则可以确定当前的抖动程度过大，此时可以进一步增大感光度，以降低拍摄时长，比如确定感光度为250。
可以理解的是，采集的图像的数量也会影响到整体的拍摄时长，采集的图像越多，拍摄时长越长，可能会导致拍摄时摄像模组的抖动程度加剧，从而影响图像质量。因此，在本申请实施例另一种可能的实现形式中，可以根据摄像模组当前的抖动程度，同时调整多帧图像的数量以及每帧图像的感光度，以使得拍摄时长控制在合适的范围内。
具体的，若摄像模组当前的抖动程度较小，则可以采集较多帧的图像，并且每帧图像的感光度可以适当压缩为较小的值，以有效抑制每帧图像的噪声、提高拍摄图像的质量；若摄像模组当前的抖动程度较大，则可以采集较少帧的图像，并且每帧图像的感光度可以适当提高为较大的值，以缩短拍摄时长。
举例来说，若确定摄像模组当前的抖动程度为“无抖动”，则可以确定当前可能为脚架拍摄模式，此时可以采集较多帧的图像，并将感光度确定为较小的值，以尽量获得更高质量的图像，比如确定多帧图像的数量为17帧，感光度为100；若确定摄像模组当前的抖动程度为“轻微抖动”，则可以确定当前可能为手持拍摄模式，此时可以采集较少帧的图像，并将感光度确定为较大的值，以降低拍摄时长，比如确定多帧图像的数量为7帧，感光度为200；若确定摄像模组当前的抖动程度为“小抖动”，则可以进一步减少多帧图像的数量，并进一步增大感光度，以降低拍摄时长，比如确定多帧图像的数量为5帧，感光度为220；若确定摄像模组当前的抖动程度为“大抖动”，则可以确定当前的抖动程度过大，此时可以 进一步减少多帧图像的数量，并进一步增大感光度，以降低拍摄时长，比如确定多帧图像的数量为3帧，感光度为250。
相应的，在根据摄像模组的抖动程度确定多帧图像的数量时，还可以预设多组曝光补偿策略，以根据多帧图像的数量确定出与多帧图像的数量相符的预设的曝光补偿策略。
需要说明的是，上述举例仅为示例性的，不能视为对本申请的限制。实际使用时，当摄像模组的抖动程度变化时，既可以同时改变待采集的图像数量和感光度，也可以改变其中之一，以获得最优的方案。其中，摄像模组的抖动程度与待采集的图像数量及每帧待采集图像对应的感光度的映射关系，可以根据实际需要预设。
在本申请实施例中，确定出多帧图像中每帧图像的目标曝光值以及每帧图像的感光度之后，即可根据每帧图像的感光度及每帧图像的目标曝光值，确定每帧图像的曝光时长。
步骤102，利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域。
其中，预设的图像识别模型，是指通过对大量经过标注后的夜景图像数据进行训练，得到的图像分割模型。
其中，目标区域，是指需要进行高强度降噪的平坦的亮度异常区域。非目标区域，是指预设帧图像中除目标区域之外的区域，一般是亮度适中且包含较多细节信息的纹理区域。即在本申请实施例一种可能的实现形式中，上述步骤102，可以包括：
利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中亮度异常的区域为目标区域。
其中，亮度异常的区域，可以是指与其周边的区域的亮度差值超过预设的阈值的区域。比如，在夜景拍摄场景中，与其他区域相比，天空区域的亮度通常很暗，因此夜景图像中的天空区域可以确定为亮度异常的区域，即目标区域；而与图像中的其他区域相比，反光区域的亮度通常很大，因此图像中的反光区域可以确定为亮度异常的区域，即目标区域；又如，光源附近的光晕区域，与光源区域相比，其亮度通常较低，因此图像中的光晕区域也可以确定为亮度异常的区域，即目标区域。
实际使用时，预设的亮度差值的阈值可以根据实际需要预设，本申请实施例对此不做限定。
需要说明的是，在夜景拍摄场景中，通过采集多帧图像并进行合成的方式，来改善夜景图像的拍摄质量时，对于天空等亮度很低的平坦区域，通常会采用过多的过曝帧的数据融合，以获得较高的亮度，而过曝帧中通常包含较大的噪声，从而造成天空等区域中包含的噪点较多。若对图像进行降噪时，采用相同的降噪参数对整幅图像进行降噪处理，容易造成天空等区域降噪效果不理想，或者其他细节信息丰富的区域中的细节信息遭到破坏。 因此，在本申请实施例中，可以通过预设的图像识别模型，从预设帧图像中确定出目标区域和非目标区域，并采用不同的降噪参数分别对目标区域与非目标区域进行降噪。
在本申请实施例中，预设的图像识别模型可以离线生成，并集成至电子设备中。训练图像识别模型时，可以首先采集大量的夜景图像数据，并对采集到的夜景图像数据的像素点按照预设的规则进行标注，以得到可以区分目标区域与非目标区域的标注信息，并对标注后的夜景图像数据通过深度学习的方式进行训练，以获得像素点的像素值与其对应的标注信息之间的规律性信息，即得到图像识别模型。训练出的图像识别模型可以对输入图像识别模型的图像数据进行分析，并根据各像素点的像素值及各像素点的像素值之间的关系，确定出每个像素点的标注信息，并对输入图像的像素点进行标注，得到可以区分目标区域与非目标区域的掩膜，从而确定出目标区域与非目标区域。
举例来说，在对采集的大量夜景图像数据进行标注时，预设的规则可以是“将目标区域对应的像素点标注为1，非目标区域对应的像素点标注为0”。对标注后的图像数据进行训练得到的图像识别模型，即可对输入图像识别模型的图像数据进行分析，并根据各像素点的像素值及各像素点的像素值之间的关系，以相同的规则对输入图像的像素点进行标注，即标注信息为“1”的像素点为目标区域，标注信息为“0”的像素点为非目标区域。
可以理解的是，将多帧图像中的预设帧图像输入预设的图像识别模型，预设的图像识别模型即可对预设帧图像进行识别处理，即根据预设帧图像中各像素点的像素值及各像素点的像素值之间的关系，确定出每个像素点的标注信息，进而确定出预设帧图像中的目标区域与非目标区域。
进一步的，由于采集的图像有多帧并且多帧图像是对齐的，因此可以从多帧图像中选取一帧图像通过预设的图像识别模型进行识别处理，并根据识别结果得到多帧图像中的目标区域与非目标区域。其中，选取的该帧图像即为预设帧图像。即在本申请实施例一种可能的实现形式中，上述步骤102之前，还可以包括：
根据所述多帧图像中每帧图像对应的曝光值，确定所述预设帧图像。
需要说明的是，在本申请实施例一种可能的实现形式中，可以将亮度和清晰度高的图像作为预设帧图像，以获得最佳的分割效果。通常来说，曝光值越大，图像的亮度和清晰度越高，因此可以将对应的曝光值最大的图像预设为预设帧图像，即可以根据每帧图像对应的曝光值确定出曝光值最大的图像，即预设帧图像。若对应的曝光值最大的图像有多帧，则可以从曝光值最大的多帧图像中随机选取出一帧作为预设帧图像。
步骤103，采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，所述目标区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
在本申请实施例中，利用预设的图像识别模型，确定出多帧图像中的目标区域及非目标区域之后，即可根据目标区域与非目标区域的特性，采用不同的降噪参数值，分别对目标区域与非目标区域进行降噪处理。
需要说明的是，目标区域中包含的噪声较大且细节信息较少，而非目标区域中包含的噪声较小且具有丰富的细节信息，而对图像进行降噪处理的同时，不可避免的会引入图像细节的模糊，这种模糊可能会抵消降噪对图像质量的提高。因此，在对多帧图像进行降噪时，可以对目标区域采用较大的降噪参数值，比如增大降噪强度，以去除目标区域的大量噪声，达到最佳的降噪效果；而对非目标区域进行降噪处理时，可以采用较小的降噪参数值，比如减小降噪强度，以在降低噪声的同时减轻降噪处理对图像细节信息的破坏，提高图像的质量。
在本申请实施例一种可能的实现形式中，可以采用空域降噪与时域降噪混合的方式对多帧图像进行降噪处理。其中，空域降噪利用单帧图像内的空间相关性进行降噪，时域降噪利用多帧图像在时间上的相关性进行降噪。空域滤波最常见的方法为低通滤波，低通滤波法通过滤除图像信号中的高频部分达到降噪的目的，但由于图像的边缘和跳跃部分也处于高频区域，容易导致图像边缘和跳跃部分出现模糊，造成集中在信号高频内的图像细节信息的损害。同时，由于空域降噪不考虑时域信息，而帧间同一位置的噪声存在随机性，容易导致降噪后相邻帧间图像内容的改变。
时域降噪最常见的方法为多图像平均法，由于图像内容在帧间具有很强的相关性，图像内容在每一帧内是连续变化的，而噪声在帧间总是随机出现，不具有相关性，噪声在每一帧内都是不连续的，因此时域降噪利用了噪声在帧间的特点，可有效的去除噪声，同时保护了图像的细节信息。但在采集多帧图像时如果出现了较大的抖动，单纯的时域降噪容易存在匹配失败或误差，出现噪声残留或“鬼影”现象。因此，采用空域降噪和时域降噪混合的方式对多帧图像进行降噪处理，可以有效的抑制多帧图像中的噪声，并且较好的保留图像的边缘、纹理等细节信息。
本申请实施例提供的夜景图像处理方法，可以根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，进而采用不同的降噪参数值，对多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，之后即可根据目标区域与非目标区域的特性，采用不同的降噪参数值分别对目标区域与非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，而且较好的保留了非目标区域的细节信息，提高了拍摄图像的质量，改善了 用户体验。
在本申请一种可能的实现形式中，还可以对目标区域进行羽化处理，确定出目标区域与非目标区域间的过渡区域，并采用不同的降噪参数值分别对目标区域、非目标区域、过渡区域进行降噪处理，以实现目降噪效果的自然过渡，进一步提高图像质量。
下面结合图2，对本申请实施例提供的另一种夜景图像处理方法进行进一步说明。
图2为本申请实施例所提供的另一种夜景图像处理方法的流程示意图。
如图2所示，该夜景图像处理方法，包括以下步骤：
步骤201，根据预设的曝光补偿模式，依次采集多帧图像。
步骤202，利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域。
上述步骤201-202的具体实现过程及原理，可以参照上述实施例的详细描述，此处不再赘述。
步骤203，对所述目标区域进行羽化处理，以确定位于所述目标区域及所述非目标区域间的过渡区域。
其中，羽化处理，是指对图像的边缘进行模糊处理，以使图像的边缘达到朦胧的效果。对目标区域进行羽化处理，即是对目标区域的边缘进行模糊处理，以达到目标区域与非目标区域之间的自然过渡的效果。
需要说明的是，可以通过调整羽化半径值控制边缘的朦胧范围，即过渡区域的大小。羽化半径值越大，朦胧范围越宽，过渡区域越大；羽化半径值越小，朦胧范围越窄，过渡区域越小。
在本申请实施例中，对目标区域进行羽化处理，即是通过一定的方法重新确定处于羽化半径内的目标区域边缘附近的像素点的像素值的过程。比如，通过均值平滑进行羽化处理，是指将某一像素点的邻域像素点的均值，重新确定为该像素点的像素值。例如，假设邻域大小是11×11，像素点A的像素值为100，其11×11的邻域中像素点的像素值均值为85，则羽化处理后像素点A的像素值为85。
步骤204，采用不同的降噪参数值，对所述多帧图像的目标区域、过渡区域及非目标区域分别进行降噪处理，以生成目标图像。
在本申请实施例中，确定出目标区域与非目标区域间的过渡区域之后，即可采用不同的降噪参数值，分别对多帧图像中的目标区域、过渡区域及非目标区域进行降噪处理，已生成目标图像。其中，目标区域对应的降噪参数值大于过渡区域对应的降噪参数值，过渡区域对应的降噪参数值大于非目标区域对应的降噪参数值。各区域对应的降噪参数值可以根据图像中包含的噪声强度确定，以获得最佳的降噪效果。
进一步的，采集的多帧图像中的噪声水平和当前的拍摄场景有关，比如拍摄时的感光度、拍摄场景的光照度、场景中的细节都可以影响采集的图像中的噪声水平。例如，感光度越大，采集的图像中的噪声水平越高；场景中的细节信息越丰富，人眼能感知的噪声越小，因此，可以根据当前的拍摄场景确定出参考降噪参数，进而根据参考降噪参数确定出各区域对应的降噪参数值。即在本申请实施例一种可能的实现形式中，上述步骤204之前，还可以包括：
根据当前的拍摄场景，确定参考降噪参数；
根据预设的各区域对应的权重值及所述参考降噪参数，确定所述目标区域、过渡区域及非目标区域当前分别对应的降噪参数值。
需要说明的是，当前的拍摄场景与采集的多帧图像中的噪声水平息息相关，因此可以根据当前的拍摄场景对多帧图像中的噪声水平进行预估，并确定出与预估的噪声水平相适应的参考降噪参数。
举例来说，若当前的感光度较小，或待拍摄的场景中的细节信息较丰富，即可以确定采集的多帧图像中的噪声水平较低，则可以将参考降噪参数确定为较小的值；若当前的感光度较大，或待拍摄的场景中的细节信息较少，即可以确定采集的多帧图像中的噪声水平较高，则可以将参考降噪参数确定为较大的值。
在本申请实施例一种可能的实现形式中，可以为各区域预设降噪时的权重值，并根据权重值与参考降噪参数确定出各区域对应的降噪参数值。比如，可以采用参考降噪参数对非目标区域进行降噪，即非目标区域对应的权重值可以是1；而目标区域对应的降噪参数值要大于非目标区域对应的降噪参数值，即目标区域对应的权重值需要大于1，比如可以是1.5；过渡区域的对应的降噪参数值需要介于目标区域对应的降噪参数值与非目标区域对应的降噪参数值之间，比如可以是1.2。
需要说明的是，上述举例仅为示例性的，不能视为对本申请的限制。实际使用时，可以根据目标区域对应的降噪参数值大于过渡区域对应的降噪参数值，过渡区域对应的降噪参数值大于非目标区域对应的降噪参数值的原则，预设各区域对应的权重值，以获得最佳的降噪效果。
本申请实施例提供的夜景图像处理方法，可以根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，之后对目标区域进行羽化处理，以确定位于目标区域及非目标区域间的过渡区域进而采用不同的降噪参数值，对多帧图像的目标区域、过渡区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于过渡区域对应的降噪参数值，过渡区域对应的降噪参数值大于非目标区域对应的 降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，并对识别出的目标区域进行羽化处理，确定出过渡区域，之后即可采用不同的降噪参数值分别对目标区域、过渡区域及非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，较好的保留了非目标区域的细节信息，而且实现了目标区域与非目标区域间降噪效果的自然过渡，进一步提高了拍摄图像的质量，改善了用户体验。
为了实现上述实施例，本申请还提出一种夜景图像处理装置。
图3为本申请实施例提供的一种夜景图像处理装置的结构示意图。
如图3所示，该夜景图像处理装置30，包括：
采集模块31，用于根据预设的曝光补偿模式，依次采集多帧图像；
第一确定模块32，用于利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域；
降噪模块33，用于采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，所述目标区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
在实际使用时，本申请实施例提供的夜景图像处理装置，可以被配置在任意电子设备中，以执行前述夜景图像处理方法。
本申请实施例提供的夜景图像处理装置，可以根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，进而采用不同的降噪参数值，对多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，之后即可根据目标区域与非目标区域的特性，采用不同的降噪参数值分别对目标区域与非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，而且较好的保留了非目标区域的细节信息，提高了拍摄图像的质量，改善了用户体验。
在本申请一种可能的实现形式中，上述夜景图像处理装置30，还包括：
第二确定模块，用于根据当前拍摄场景的光照度和摄像模组当前的抖动程度，确定所述预设的曝光补偿模式。
进一步的，在本申请另一种可能的实现形式中，上述第二确定模块，具体用于：
根据当前拍摄场景的光照度，确定所述多帧图像中每帧图像的目标曝光值；
根据所述摄像模组当前的抖动程度，确定所述每帧图像的感光度；
根据所述每帧图像的感光度及每帧图像的目标曝光值，确定所述每帧图像的曝光时长。
进一步的，在本申请再一种可能的实现形式中，上述第二确定模块，还用于：
根据所述当前拍摄场景的光照度，确定基准曝光值；
根据所述基准曝光值及预设的曝光补偿策略，确定所述多帧图像中每帧图像的目标曝光值。
进一步的，在本申请又一种可能的实现形式中，上述夜景图像处理装置30，还包括：
第三确定模块，用于根据所述多帧图像中每帧图像对应的曝光值，确定所述预设帧图像。
进一步的，在本申请又一种可能的实现形式中，上述夜景图像处理装置30，还包括：
第四确定模块，用于对所述目标区域进行羽化处理，以确定位于所述目标区域及所述非目标区域间的过渡区域。
相应的，上述降噪模块33，具体用于：
采用不同的降噪参数值，对所述多帧图像的目标区域、过渡区域及非目标区域分别进行降噪处理。
进一步的，在本申请另一种可能的实现形式中，上述目标区域对应的降噪参数值大于上述过渡区域对应的降噪参数值，上述过渡区域对应的降噪参数值大于上述非目标区域对应的降噪参数值。
进一步的，在本申请再一种可能的实现形式中，上述夜景图像处理装置30，还包括：
第五确定模块，用于根据当前的拍摄场景，确定参考降噪参数；
第六确定模块，用于根据预设的各区域对应的权重值及所述参考降噪参数，确定所述目标区域及非目标区域当前分别对应的降噪参数值。
进一步的，在本申请又一种可能的实现形式中，上述第一确定模块32，具体用于：
利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中亮度异常的区域为目标区域。
需要说明的是，前述对图1、图2所示的夜景图像处理方法实施例的解释说明也适用于该实施例的夜景图像处理装置30，此处不再赘述。
本申请实施例提供的夜景图像处理装置，可以根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，之后对目标区域进行羽化处理，以确定位于目标区域及非目标区域间的过渡区域进而采用不同的降噪参数值，对多帧图像的目标区域、过渡区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于过渡区域对应的降噪参数值，过渡区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，并对识 别出的目标区域进行羽化处理，确定出过渡区域，之后即可采用不同的降噪参数值分别对目标区域、过渡区域及非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，较好的保留了非目标区域的细节信息，而且实现了目标区域与非目标区域间降噪效果的自然过渡，进一步提高了拍摄图像的质量，改善了用户体验。
为了实现上述实施例，本申请还提出一种电子设备。
图4为本申请实施例提供的电子设备的结构示意图。
如图4所示，上述电子设备200包括：摄像模组201、存储器210、处理器220及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时，实现本申请实施例所述的夜景图像处理方法。
如图5所示，本申请实施例提供的电子设备200，还可以包括：
存储器210及处理器220，连接不同组件(包括存储器210和处理器220)的总线230，存储器210存储有计算机程序，当处理器220执行所述程序时实现本申请实施例所述的夜景图像处理方法。
总线230表示几类总线结构中的一种或多种，包括存储器总线或者存储器控制器，外围总线，图形加速端口，处理器或者使用多种总线结构中的任意总线结构的局域总线。举例来说，这些体系结构包括但不限于工业标准体系结构(ISA)总线，微通道体系结构(MAC)总线，增强型ISA总线、视频电子标准协会(VESA)局域总线以及外围组件互连(PCI)总线。
电子设备200典型地包括多种电子设备可读介质。这些介质可以是任何能够被电子设备200访问的可用介质，包括易失性和非易失性介质，可移动的和不可移动的介质。
存储器210还可以包括易失性存储器形式的计算机系统可读介质，例如随机存取存储器(RAM)240和/或高速缓存存储器250。电子设备200可以进一步包括其它可移动/不可移动的、易失性/非易失性计算机系统存储介质。仅作为举例，存储系统260可以用于读写不可移动的、非易失性磁介质(图5未显示，通常称为“硬盘驱动器”)。尽管图5中未示出，可以提供用于对可移动非易失性磁盘(例如“软盘”)读写的磁盘驱动器，以及对可移动非易失性光盘(例如CD-ROM,DVD-ROM或者其它光介质)读写的光盘驱动器。在这些情况下，每个驱动器可以通过一个或者多个数据介质接口与总线230相连。存储器210可以包括至少一个程序产品，该程序产品具有一组(例如至少一个)程序模块，这些程序模块被配置以执行本申请各实施例的功能。
具有一组(至少一个)程序模块270的程序/实用工具280，可以存储在例如存储器210中，这样的程序模块270包括——但不限于——操作系统、一个或者多个应用程序、其它程序模块以及程序数据，这些示例中的每一个或某种组合中可能包括网络环境的实现。程 序模块270通常执行本申请所描述的实施例中的功能和/或方法。
电子设备200也可以与一个或多个外部设备290(例如键盘、指向设备、显示器291等)通信，还可与一个或者多个使得用户能与该电子设备200交互的设备通信，和/或与使得该电子设备200能与一个或多个其它计算设备进行通信的任何设备(例如网卡，调制解调器等等)通信。这种通信可以通过输入/输出(I/O)接口292进行。并且，电子设备200还可以通过网络适配器293与一个或者多个网络(例如局域网(LAN)，广域网(WAN)和/或公共网络，例如因特网)通信。如图所示，网络适配器293通过总线230与电子设备200的其它模块通信。应当明白，尽管图中未示出，可以结合电子设备200使用其它硬件和/或软件模块，包括但不限于：微代码、设备驱动器、冗余处理单元、外部磁盘驱动阵列、RAID系统、磁带驱动器以及数据备份存储系统等。
处理器220通过运行存储在存储器210中的程序，从而执行各种功能应用以及数据处理。
需要说明的是，本实施例的电子设备的实施过程和技术原理参见前述对本申请实施例的夜景图像处理方法的解释说明，此处不再赘述。
本申请实施例提供的电子设备，可以执行如前所述的夜景图像处理方法，根据预设的曝光补偿模式，依次采集多帧图像，并利用预设的图像识别模型，对多帧图像中的预设帧图像进行识别处理，以确定预设帧图像中的目标区域及非目标区域，进而采用不同的降噪参数值，对多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，目标区域对应的降噪参数值大于非目标区域对应的降噪参数值。由此，通过利用预设的图像识别模型，对预设帧图像进行分割处理，之后即可根据目标区域与非目标区域的特性，采用不同的降噪参数值分别对目标区域与非目标区域进行降噪处理，从而不仅保证了图像整体的降噪效果和目标区域的纯净度，而且较好的保留了非目标区域的细节信息，提高了拍摄图像的质量，改善了用户体验。
为了实现上述实施例，本申请还提出一种计算机可读存储介质。
其中，该计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时，以实现本申请实施例所述的夜景图像处理方法。
为了实现上述实施例，本申请再一方面实施例提供一种计算机程序，该程序被处理器执行时，以实现本申请实施例所述的夜景图像处理方法。
一种可选实现形式中，本实施例可以采用一个或多个计算机可读的介质的任意组合。计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质。计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、装置或器件，或者任意以上的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括： 具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本文件中，计算机可读存储介质可以是任何包含或存储程序的有形介质，该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。
计算机可读的信号介质可以包括在基带中或者作为载波一部分传播的数据信号，其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式，包括——但不限于——电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质，该计算机可读介质可以发送、传播或者传输用于由指令执行系统、装置或者器件使用或者与其结合使用的程序。
计算机可读介质上包含的程序代码可以用任何适当的介质传输，包括——但不限于——无线、电线、光缆、RF等等，或者上述的任意合适的组合。
可以以一种或多种程序设计语言或其组合来编写用于执行本发明操作的计算机程序代码，所述程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++，还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言。程序代码可以完全地在用户电子设备上执行、部分地在用户电子设备上执行、作为一个独立的软件包执行、部分在用户电子设备上部分在远程电子设备上执行、或者完全在远程电子设备或服务器上执行。在涉及远程电子设备的情形中，远程电子设备可以通过任意种类的网络——包括局域网(LAN)或广域网(WAN)—连接到用户电子设备，或者，可以连接到外部电子设备(例如利用因特网服务提供商来通过因特网连接)。
本领域技术人员在考虑说明书及实践这里申请的发明后，将容易想到本申请的其它实施方案。本申请旨在涵盖本申请的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本申请的一般性原理并包括本申请未发明的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本申请的真正范围和精神由权利要求指出。
应当理解的是，本申请并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本申请的范围仅由所附的权利要求来限制。

Claims (20)

  1. 一种夜景图像处理方法，其特征在于，包括：
    根据预设的曝光补偿模式，依次采集多帧图像；
    利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域；
    采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，所述目标区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
  2. 如权利要求1所述的方法，其特征在于，所述根据预设的曝光补偿模式，依次采集多帧图像之前，还包括：
    根据当前拍摄场景的光照度和摄像模组当前的抖动程度，确定所述预设的曝光补偿模式。
  3. 如权利要求2所述的方法，其特征在于，所述根据当前拍摄场景的光照度和摄像模组当前的抖动程度，确定所述预设的曝光补偿模式，包括：
    根据当前拍摄场景的光照度，确定所述多帧图像中每帧图像的目标曝光值；
    根据所述摄像模组当前的抖动程度，确定所述每帧图像的感光度；
    根据所述每帧图像的感光度及每帧图像的目标曝光值，确定所述每帧图像的曝光时长。
  4. 如权利要求3所述的方法，其特征在于，所述根据当前拍摄场景的光照度，确定所述多帧图像中每帧图像的目标曝光值，包括：根据所述当前拍摄场景的光照度，确定基准曝光值；
    根据所述基准曝光值及预设的曝光补偿策略，确定所述多帧图像中每帧图像的目标曝光值。
  5. 如权利要求1-4任一项所述的方法，其特征在于，所述对所述多帧图像中的预设帧图像进行识别处理之前，还包括:
    根据所述多帧图像中每帧图像对应的曝光值，确定所述预设帧图像。
  6. 如权利要求1-5任一项所述的方法，其特征在于，所述确定所述预设帧图像中的目标区域及非目标区域之后，还包括：
    对所述目标区域进行羽化处理，以确定位于所述目标区域及所述非目标区域间的过渡区域；
    所述采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，包括：
    采用不同的降噪参数值，对所述多帧图像的目标区域、过渡区域及非目标区域分别进行降噪处理。
  7. 如权利要求6所述的方法，其特征在于，所述目标区域对应的降噪参数值大于所述过渡区域对应的降噪参数值，所述过渡区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
  8. 如权利要求1-7任一项所述的方法，其特征在于，所述采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理之前，还包括：
    根据当前的拍摄场景，确定参考降噪参数；
    根据预设的各区域对应的权重值及所述参考降噪参数，确定所述目标区域及非目标区域当前分别对应的降噪参数值。
  9. 如权利要求1-8任一项所述的方法，其特征在于，所述确定所述预设帧图像中的目标区域，包括：
    利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中亮度异常的区域为目标区域。
  10. 一种夜景图像处理装置，其特征在于，包括：
    采集模块，用于根据预设的曝光补偿模式，依次采集多帧图像；
    第一确定模块，用于利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中的目标区域及非目标区域；
    降噪模块，用于采用不同的降噪参数值，对所述多帧图像的目标区域及非目标区域分别进行降噪处理，以生成目标图像，其中，所述目标区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
  11. 如权利要求10所述的装置，其特征在于，还包括：
    第二确定模块，用于根据当前拍摄场景的光照度和摄像模组当前的抖动程度，确定所述预设的曝光补偿模式。
  12. 如权利要求11所述的装置，其特征在于，所述第二确定模块，具体用于：
    根据当前拍摄场景的光照度，确定所述多帧图像中每帧图像的目标曝光值；
    根据所述摄像模组当前的抖动程度，确定所述每帧图像的感光度；
    根据所述每帧图像的感光度及每帧图像的目标曝光值，确定所述每帧图像的曝光时长。
  13. 如权利要求12所述的装置，其特征在于，所述第二确定模块，还用于：
    根据所述当前拍摄场景的光照度，确定基准曝光值；
    根据所述基准曝光值及预设的曝光补偿策略，确定所述多帧图像中每帧图像的目标曝光值。
  14. 如权利要求10-13任一项所述的装置，其特征在于，还包括：
    第三确定模块，用于根据所述多帧图像中每帧图像对应的曝光值，确定所述预设帧图像。
  15. 如权利要求10-14任一项所述的装置，其特征在于，还包括：
    第四确定模块，用于对所述目标区域进行羽化处理，以确定位于所述目标区域及所述非目标区域间的过渡区域；
    所述降噪模块，具体用于采用不同的降噪参数值，对所述多帧图像的目标区域、过渡区域及非目标区域分别进行降噪处理。
  16. 如权利要求15所述的装置，其特征在于，所述目标区域对应的降噪参数值大于所述过渡区域对应的降噪参数值，所述过渡区域对应的降噪参数值大于所述非目标区域对应的降噪参数值。
  17. 如权利要求10-16任一项所述的装置，其特征在于，还包括：
    第五确定模块，用于根据当前的拍摄场景，确定参考降噪参数；
    第六确定模块，用于根据预设的各区域对应的权重值及所述参考降噪参数，确定所述目标区域及非目标区域当前分别对应的降噪参数值。
  18. 如权利要求10-17任一项所述的装置，其特征在于，所述第一确定模块，具体用于：
    利用预设的图像识别模型，对所述多帧图像中的预设帧图像进行识别处理，以确定所述预设帧图像中亮度异常的区域为目标区域。
  19. 一种电子设备，其特征在于，包括：所述摄影模组、存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述计算机程序时，实现如权利要求1-9中任一项所述的夜景图像处理方法。
  20. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现如权利要求1-9中任一项所述的夜景图像处理方法。
PCT/CN2019/101430 2018-11-22 2019-08-19 夜景图像处理方法、装置、电子设备及存储介质 WO2020103503A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811399541.0A CN109348089B (zh) 2018-11-22 2018-11-22 夜景图像处理方法、装置、电子设备及存储介质
CN201811399541.0 2018-11-22

Publications (1)

Publication Number Publication Date
WO2020103503A1 true WO2020103503A1 (zh) 2020-05-28

Family

ID=65317460

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/101430 WO2020103503A1 (zh) 2018-11-22 2019-08-19 夜景图像处理方法、装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN109348089B (zh)
WO (1) WO2020103503A1 (zh)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112418322A (zh) * 2020-11-24 2021-02-26 苏州爱医斯坦智能科技有限公司 影像数据处理方法、装置、电子设备及存储介质
CN112785535A (zh) * 2020-12-30 2021-05-11 北京迈格威科技有限公司 一种获取夜景光轨图像的方法、装置以及可手持终端
CN112987008A (zh) * 2021-02-09 2021-06-18 上海眼控科技股份有限公司 一种相对深度测量方法、装置、设备及存储介质
CN114511450A (zh) * 2020-11-17 2022-05-17 北京小米移动软件有限公司 图像降噪方法、图像降噪装置、终端及存储介质
CN114697469A (zh) * 2022-03-15 2022-07-01 华能大理风力发电有限公司洱源分公司 适用于光伏电站的视频处理方法、装置及电子设备
CN114783454A (zh) * 2022-04-27 2022-07-22 北京百度网讯科技有限公司 一种模型训练、音频降噪方法、装置、设备及存储介质
CN115334250A (zh) * 2022-08-09 2022-11-11 阿波罗智能技术(北京)有限公司 一种图像处理方法、装置及电子设备
CN115379130A (zh) * 2022-08-25 2022-11-22 上海联影医疗科技股份有限公司 一种自动曝光控制系统、方法、装置和存储介质
WO2023077938A1 (zh) * 2021-11-05 2023-05-11 荣耀终端有限公司 生成视频帧的方法、装置、电子设备及存储介质
CN116977411A (zh) * 2022-12-01 2023-10-31 开立生物医疗科技(武汉)有限公司 内镜移动速度估计方法及装置、电子设备及存储介质

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109348089B (zh) * 2018-11-22 2020-05-22 Oppo广东移动通信有限公司 夜景图像处理方法、装置、电子设备及存储介质
CN109873953A (zh) * 2019-03-06 2019-06-11 深圳市道通智能航空技术有限公司 图像处理方法、夜间拍摄方法、图像处理芯片及航拍相机
CN110136085B (zh) * 2019-05-17 2022-03-29 凌云光技术股份有限公司 一种图像的降噪方法及装置
CN110264473B (zh) * 2019-06-13 2022-01-04 Oppo广东移动通信有限公司 基于多帧图像的图像处理方法、装置及电子设备
CN110246101B (zh) * 2019-06-13 2023-03-28 Oppo广东移动通信有限公司 图像处理方法和装置
CN110443766B (zh) * 2019-08-06 2022-05-31 厦门美图之家科技有限公司 图像处理方法、装置、电子设备及可读存储介质
CN112651909B (zh) * 2019-10-10 2024-03-15 北京字节跳动网络技术有限公司 图像合成方法、装置、电子设备及计算机可读存储介质
CN111191498A (zh) * 2019-11-07 2020-05-22 腾讯科技（深圳）有限公司 行为识别方法和相关产品
CN112907454B (zh) * 2019-11-19 2023-08-08 杭州海康威视数字技术股份有限公司 获取图像的方法、装置、计算机设备和存储介质
CN111028189B (zh) * 2019-12-09 2023-06-27 Oppo广东移动通信有限公司 图像处理方法、装置、存储介质及电子设备
CN111147695B (zh) * 2019-12-31 2022-05-13 Oppo广东移动通信有限公司 图像处理方法、图像处理器、拍摄装置和电子设备
CN111246091B (zh) * 2020-01-16 2021-09-03 北京迈格威科技有限公司 一种动态自动曝光控制方法和装置及电子设备
CN111583144B (zh) * 2020-04-30 2023-08-25 深圳市商汤智能传感科技有限公司 图像降噪方法及装置、电子设备和存储介质
CN114928694A (zh) * 2022-04-25 2022-08-19 深圳市慧鲤科技有限公司 图像获取方法和装置、设备、介质

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120044357A1 (en) * 2010-08-19 2012-02-23 Samsung Electro-Mechanics Co., Ltd. Image processing apparatus and method for night vision
CN105608676A (zh) * 2015-12-23 2016-05-25 浙江宇视科技有限公司 一种视频图像的增强方法及装置
CN105681665A (zh) * 2016-02-29 2016-06-15 广东欧珀移动通信有限公司 控制方法、控制装置及电子装置
CN106303250A (zh) * 2016-08-26 2017-01-04 维沃移动通信有限公司 一种图像处理方法及移动终端
CN106412385A (zh) * 2016-10-17 2017-02-15 湖南国科微电子股份有限公司 一种视频图像3d降噪方法及装置
CN106875347A (zh) * 2016-12-30 2017-06-20 努比亚技术有限公司 一种图片处理装置及方法
CN108629747A (zh) * 2018-04-25 2018-10-09 腾讯科技（深圳）有限公司 图像增强方法、装置、电子设备及存储介质
CN108683859A (zh) * 2018-08-16 2018-10-19 Oppo广东移动通信有限公司 拍照优化方法、装置、存储介质及终端设备
CN109348089A (zh) * 2018-11-22 2019-02-15 Oppo广东移动通信有限公司 夜景图像处理方法、装置、电子设备及存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7065255B2 (en) * 2002-05-06 2006-06-20 Eastman Kodak Company Method and apparatus for enhancing digital images utilizing non-image data
US8023733B2 (en) * 2006-06-08 2011-09-20 Panasonic Corporation Image processing device, image processing method, image processing program, and integrated circuit
US8355059B2 (en) * 2009-02-06 2013-01-15 Canon Kabushiki Kaisha Image capturing apparatus and control method thereof
CN105208281B (zh) * 2015-10-09 2019-12-03 Oppo广东移动通信有限公司 一种夜景拍摄方法及装置
CN105812618B (zh) * 2016-03-17 2019-09-17 浙江大华技术股份有限公司 一种运动检测方法及运动检测装置
CN106375676A (zh) * 2016-09-20 2017-02-01 广东欧珀移动通信有限公司 终端设备的拍照控制方法、装置和终端设备
CN107770438B (zh) * 2017-09-27 2019-11-29 维沃移动通信有限公司 一种拍照方法及移动终端
CN107948538B (zh) * 2017-11-14 2020-02-21 Oppo广东移动通信有限公司 成像方法、装置、移动终端和存储介质
CN108111762A (zh) * 2017-12-27 2018-06-01 努比亚技术有限公司 一种图像处理方法、终端及计算机可读存储介质
CN108833804A (zh) * 2018-09-20 2018-11-16 Oppo广东移动通信有限公司 成像方法、装置和电子设备

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120044357A1 (en) * 2010-08-19 2012-02-23 Samsung Electro-Mechanics Co., Ltd. Image processing apparatus and method for night vision
CN105608676A (zh) * 2015-12-23 2016-05-25 浙江宇视科技有限公司 一种视频图像的增强方法及装置
CN105681665A (zh) * 2016-02-29 2016-06-15 广东欧珀移动通信有限公司 控制方法、控制装置及电子装置
CN106303250A (zh) * 2016-08-26 2017-01-04 维沃移动通信有限公司 一种图像处理方法及移动终端
CN106412385A (zh) * 2016-10-17 2017-02-15 湖南国科微电子股份有限公司 一种视频图像3d降噪方法及装置
CN106875347A (zh) * 2016-12-30 2017-06-20 努比亚技术有限公司 一种图片处理装置及方法
CN108629747A (zh) * 2018-04-25 2018-10-09 腾讯科技（深圳）有限公司 图像增强方法、装置、电子设备及存储介质
CN108683859A (zh) * 2018-08-16 2018-10-19 Oppo广东移动通信有限公司 拍照优化方法、装置、存储介质及终端设备
CN109348089A (zh) * 2018-11-22 2019-02-15 Oppo广东移动通信有限公司 夜景图像处理方法、装置、电子设备及存储介质

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114511450A (zh) * 2020-11-17 2022-05-17 北京小米移动软件有限公司 图像降噪方法、图像降噪装置、终端及存储介质
CN112418322A (zh) * 2020-11-24 2021-02-26 苏州爱医斯坦智能科技有限公司 影像数据处理方法、装置、电子设备及存储介质
CN112785535A (zh) * 2020-12-30 2021-05-11 北京迈格威科技有限公司 一种获取夜景光轨图像的方法、装置以及可手持终端
CN112987008A (zh) * 2021-02-09 2021-06-18 上海眼控科技股份有限公司 一种相对深度测量方法、装置、设备及存储介质
WO2023077938A1 (zh) * 2021-11-05 2023-05-11 荣耀终端有限公司 生成视频帧的方法、装置、电子设备及存储介质
CN114697469B (zh) * 2022-03-15 2024-02-06 华能大理风力发电有限公司洱源分公司 适用于光伏电站的视频处理方法、装置及电子设备
CN114697469A (zh) * 2022-03-15 2022-07-01 华能大理风力发电有限公司洱源分公司 适用于光伏电站的视频处理方法、装置及电子设备
CN114783454A (zh) * 2022-04-27 2022-07-22 北京百度网讯科技有限公司 一种模型训练、音频降噪方法、装置、设备及存储介质
CN114783454B (zh) * 2022-04-27 2024-06-04 北京百度网讯科技有限公司 一种模型训练、音频降噪方法、装置、设备及存储介质
CN115334250A (zh) * 2022-08-09 2022-11-11 阿波罗智能技术(北京)有限公司 一种图像处理方法、装置及电子设备
CN115334250B (zh) * 2022-08-09 2024-03-08 阿波罗智能技术(北京)有限公司 一种图像处理方法、装置及电子设备
CN115379130B (zh) * 2022-08-25 2024-03-29 上海联影医疗科技股份有限公司 一种自动曝光控制系统、方法、装置和存储介质
CN115379130A (zh) * 2022-08-25 2022-11-22 上海联影医疗科技股份有限公司 一种自动曝光控制系统、方法、装置和存储介质
CN116977411A (zh) * 2022-12-01 2023-10-31 开立生物医疗科技(武汉)有限公司 内镜移动速度估计方法及装置、电子设备及存储介质
CN116977411B (zh) * 2022-12-01 2024-03-19 开立生物医疗科技(武汉)有限公司 内镜移动速度估计方法及装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN109348089B (zh) 2020-05-22
CN109348089A (zh) 2019-02-15

Similar Documents

Publication Publication Date Title
WO2020103503A1 (zh) 夜景图像处理方法、装置、电子设备及存储介质
KR102642993B1 (ko) 야경 촬영 방법, 장치, 전자설비, 및 저장 매체
US11582400B2 (en) Method of image processing based on plurality of frames of images, electronic device, and storage medium
WO2020057198A1 (zh) 图像处理方法、装置、电子设备及存储介质
WO2020207262A1 (zh) 基于多帧图像的图像处理方法、装置、电子设备
CN110062160B (zh) 图像处理方法和装置
US11532076B2 (en) Image processing method, electronic device and storage medium
WO2021082580A1 (zh) 一种夜景高动态范围图像生成方法、装置和电子设备
WO2020038087A1 (zh) 超级夜景模式下的拍摄控制方法、装置和电子设备
CN108322646B (zh) 图像处理方法、装置、存储介质及电子设备
WO2020038074A1 (zh) 曝光控制方法、装置以及电子设备
CN109218627B (zh) 图像处理方法、装置、电子设备及存储介质
WO2020038072A1 (zh) 曝光控制方法、装置和电子设备
WO2020034737A1 (zh) 成像控制方法、装置、电子设备以及计算机可读存储介质
WO2020207261A1 (zh) 基于多帧图像的图像处理方法、装置、电子设备
CN110191291B (zh) 基于多帧图像的图像处理方法和装置
WO2019105297A1 (zh) 图像虚化处理方法、装置、移动设备及存储介质
US11233948B2 (en) Exposure control method and device, and electronic device
CN109919116B (zh) 场景识别方法、装置、电子设备及存储介质
CN109618102B (zh) 对焦处理方法、装置、电子设备及存储介质
CN110264420B (zh) 基于多帧图像的图像处理方法和装置
CN109361853B (zh) 图像处理方法、装置、电子设备及存储介质
CN113298735A (zh) 图像处理方法、装置、电子设备及存储介质
CN110276730B (zh) 图像处理方法、装置、电子设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19887040

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19887040

Country of ref document: EP

Kind code of ref document: A1