WO2018166170A1

WO2018166170A1 - Image processing method and device, and intelligent conferencing terminal

Info

Publication number: WO2018166170A1
Application number: PCT/CN2017/103282
Authority: WO
Inventors: 运如靖
Original assignee: 广州视源电子科技股份有限公司; 广州视睿电子科技有限公司
Priority date: 2017-03-17
Filing date: 2017-09-25
Publication date: 2018-09-20
Also published as: CN106803920A; CN106803920B

Abstract

Disclosed in embodiments of the present invention are an image processing method and device, and an intelligent conferencing terminal. The method comprises: obtaining a current real scene image frame captured by a camera, and determining a target focus image in the current real scene image frame; determining a far limit of depth of field of the current real scene image frame according to the target focus image; and adjusting image parameter information of an image area corresponding to the far limit of depth of field. By means of the method, a local image in an image frame captured during a video call can be adjusted, a target area to be processed can be determined and processed efficiently, the flexibility of image processing is greatly enhanced, and display effects of video participants on intelligent terminals are effectively improved.

Description

Image processing method, device and intelligent conference terminal

Technical field

The present invention relates to the field of image processing technologies, and in particular, to a method, an apparatus, and an intelligent conference terminal for image processing.

Background technique

At present, a smart terminal usually has a video call function, and after the smart terminal establishes a connection with other smart terminals, it can perform a video call based on the video call function it has.

Generally, during a video call, the smart terminal captures the target object in real time through the camera to form an image frame, and continuously transmits the captured image frame to other intelligent terminal devices. For a large intelligent terminal with a video call function, such as a smart conference tablet, the terminal itself is often fixed and generally disposed at a position opposite to the window, and the user participating in the video is based on the smart terminal performing a video call. It is often in a state of backlight. At this time, the image information of the user cannot be clearly displayed in the image frame captured by the camera on the smart terminal device, and the closer the user is located to the window, the less clear the user image information displayed in the image frame is. Thus, the image information in the image frame needs to be processed before the image frame is sent to other smart terminal devices.

In the prior art, when image information is processed, the processing of the entire image is often performed, and the processing manner thereof has limitations.

Summary of the invention

Embodiments of the present invention provide a method, an apparatus, and an intelligent conference terminal for image processing, which increase the flexibility of image processing, thereby achieving the purpose of clearly displaying a target object in a captured image frame during a video call.

In one aspect, an embodiment of the present invention provides a method for image processing, including:

Obtaining a current live image frame captured by the camera and determining a target focused image in the current live image frame;

Determining a depth of field limit value of the current live image frame according to the target focused image;

The image region information is subjected to adjustment processing on the image region corresponding to the depth of field limit value.

In another aspect, an embodiment of the present invention provides an apparatus for image processing, including:

a real image acquisition module, configured to acquire a current live image frame captured by the camera;

a focused image determining module, configured to determine a target focused image in the current live image frame;

a depth of field limit determining module, configured to determine a depth of field limit value of the current live image frame according to the target focused image;

The image parameter adjustment module is configured to perform image parameter information adjustment processing on the image region corresponding to the depth of field limit value.

In another aspect, an embodiment of the present invention provides an intelligent conference terminal, including: at least two cameras having optical axes parallel, and an apparatus for image processing according to the foregoing embodiment of the present invention.

In the above method, device and intelligent conference terminal for image processing, first, a current live image frame captured by a camera is acquired, and a target focused image in a current live image frame is determined; and then a depth of field of the current live image frame is determined according to the target focused image. The threshold value is finally adjusted for the image parameter information corresponding to the image region corresponding to the depth of field limit value. The above method, device and intelligent conference terminal can adjust the partial image in the image frame captured during the video call, and efficiently realize the determination and processing of the target area to be processed, thereby further increasing the flexibility of image processing. Effectively enhance the display effect of video participants on the smart terminal.

DRAWINGS

FIG. 1 is a schematic flowchart diagram of a method for image processing according to Embodiment 1 of the present invention;

2 is a schematic flowchart of a method for image processing according to Embodiment 2 of the present invention;

FIG. 3 is a structural block diagram of an apparatus for image processing according to Embodiment 3 of the present invention.

detailed description

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. It should also be noted that, for ease of description, only some, but not all, of the structures related to the present invention are shown in the drawings.

Embodiment 1

FIG. 1 is a schematic flowchart of a method for image processing according to a first embodiment of the present invention. The method is applicable to image processing of a captured image frame during a video call, and the method may be performed by an image processing device, where The device can be implemented by software and/or hardware and is generally integrated on a smart terminal having a video call function.

In this embodiment, the smart terminal may be a smart mobile terminal such as a mobile phone, a tablet computer, or a notebook, or a fixed electronic device with a video call function such as a desktop computer or a smart conference terminal.

In this embodiment, the application scenario is preferably a video call. For a fixed smart terminal, if the fixed camera is corresponding to the window in the room, and the light intensity of the outdoor environment is greater than the light intensity of the indoor environment, Then the video participant in the current live image frame captured by the camera will be in backlight The status may not be clearly displayed in the current live image frame. Therefore, according to the image processing method provided in this embodiment, the specific image area where the indoor window is located is determined, thereby adjusting the image parameters such as image brightness and image sharpness of the image area where the indoor window is located.

As shown in FIG. 1, a method for image processing according to Embodiment 1 of the present invention includes the following operations:

S101. Acquire a current live image frame captured by the camera, and determine a target focused image in the current live image frame.

In this embodiment, when a video call is made, the image of the capture space can be captured by the camera in real time, thereby forming a current live image frame. In addition, when capturing an image in the capture space, one subject is selected as the target focused image. In the embodiment, when the image capture is performed, the dynamic subject in the capture space can be used as the target focused image. The image area corresponding to the dynamic subject is determined in the current real image frame, and the image corresponding to the preset pixel information may be used as the target focused image. In this case, the preset pixel information needs to be in the current real scene. The corresponding image area in the image frame is determined to be the target focused image.

S102. Determine a depth of field limit value of the current live image frame according to the target focused image.

In this embodiment, according to the target focused image determined by the above steps, the actual distance of the target focused image to the front node of the camera may be determined, and the actual distance is equivalent to the focus distance of the camera at this time. In this embodiment, the focus is The distance may be determined according to current pixel information of the image focused image and corresponding depth of field information, and further, the depth of field range of the image frame captured by the camera may be determined according to the focus distance and the attribute parameter of the camera.

Generally, the depth of field range is formed by a depth of field near threshold value and a depth of field limit value, and the depth of field near threshold value can display the closest distance between the image in the current live image frame and the camera; the depth of field limit value may specifically Think of the farthest between the image and the camera that can be displayed in the current live image frame The distance, therefore, determines the depth of field limit value of the current live image frame based on the determined depth of field range.

S103: Perform image coordinate information adjustment processing on the image region corresponding to the depth of field limit value.

In this step, the current real image frame can be understood as an image frame having depth of field information. After determining the depth of field limit value, the image region corresponding to the far depth limit value of the depth of field can be determined in the current real image frame, and then determined. The image area is subjected to mediation processing based on its image parameter information.

Illustratively, for a stationary smart terminal, when the fixedly placed camera is opposite to the window in the room and a video call is made, the video participant is reduced in order to reduce the light intensity of the indoor window in the captured current live picture frame. The effect of the display screen can be regarded as the area where the indoor window is located by using the image area corresponding to the far depth limit value, thereby locally adjusting the determined image area, thereby achieving the purpose of clearly displaying the video participants.

A method for image processing according to Embodiment 1 of the present invention first acquires a current live image frame captured by a camera, and determines a target focused image in a current live image frame; and then determines a current live image frame according to the target focused image. The depth of field is far from the limit value; finally, the image parameter information is adjusted to the image area corresponding to the depth of field limit value. By using the method, the partial image in the image frame captured during the video call can be adjusted, thereby achieving the purpose of clearly displaying the video participant, and the flexibility of image processing is better.

Embodiment 2

FIG. 2 is a schematic flowchart diagram of a method for image processing according to Embodiment 2 of the present invention. The embodiment of the present invention is optimized based on the foregoing embodiment. In this embodiment, the current real-life image frame captured by the camera is further optimized to be: captured by at least two cameras respectively. a pre-image frame; performing image synthesizing processing on the at least two current image frames respectively captured to obtain a current real-image frame; wherein each pixel in the current live image frame has corresponding depth-of-field information.

On the basis of the above optimization, the target focused image in the current real image frame is also determined to be: determining a subject in the current live image frame according to the character image feature, and determining to form the taken person Current pixel information; determining whether the subject is present in the acquired previous live image frame; if the subject is present, determining the composition of the subject in the previous live image frame Historical pixel information, and determining whether the current pixel information matches the historical pixel information, and if not, determining that the position of the subject is changed, determining the subject to be a target focused image; if yes, Determining average pixel information according to current pixel information of each of the captured persons, determining an area corresponding to the average pixel information as a target focused image; if the captured person does not exist, acquiring preset focused pixel information, The corresponding area of the focused pixel information in the current live image frame is determined as the target focused image.

Further, determining the depth of field limit value of the current live image frame according to the target focused image may be optimized to: determine, according to current pixel information of the target focused image in the current live image frame, a plane coordinate information of the target focused image; determining a depth value of the target focused image according to the depth information corresponding to the current pixel information; determining the target focused image to the camera according to the plane coordinate information and the depth value The actual focus distance; determining the depth of field limit value of the current live image frame according to the actual focus distance and the acquired camera attribute parameter.

In addition, the present embodiment further performs an adjustment process of the image parameter information on the image region corresponding to the depth of field limit value, and is specifically optimized to: acquire image parameter information of the image region corresponding to the depth of field limit value, the image parameter information. The method includes: an image RGB ratio, a color contrast, and an image sharpness; and when the image parameter information does not meet the set standard parameter information, controlling to adjust the image region Image brightness, color contrast, and/or image sharpness are such that the image parameter information conforms to the standard parameter information.

As shown in FIG. 2, a second embodiment of the present invention provides a method for image processing, which specifically includes the following operations:

S201. Acquire a current image frame respectively captured by at least two cameras.

Generally, in order to acquire the depth of field information of the captured image frame, it is necessary to capture an image frame having a stereoscopic space, whereby at least two cameras arranged in parallel with the optical axis can perform image capturing from different angles in real time.

In the embodiment, when the image capturing is performed by using a plurality of cameras, the setting positions of the at least two cameras used on the smart terminal are different, and for the same subject, the image captured by the object in different cameras is different. The pixel positions in the frame are different, and the depth of field information of the object can be determined according to different pixel position information.

S202. Perform image synthesis processing on at least two current image frames respectively captured to obtain a current real image frame.

In this step, the current image frames captured by different cameras can be combined to obtain a current live image frame having a stereoscopic sense. It can be understood that each pixel in the synthesized current real image frame has corresponding depth of field information. Specifically, the process of determining the depth information of each pixel may be described as: stereo matching the current image frame captured by different cameras. Therefore, the disparity values of the same corresponding points in different current image frames are obtained, and then the depth information of different pixel points can be determined according to the relationship between the disparity values and the depths.

In this embodiment, the depth information of each pixel in the current live image frame may be stored for selection of a subsequent image area to be processed.

S203. Determine a subject in the current live image frame according to the character image feature, and determine current pixel information that constitutes the subject.

In the present embodiment, steps S203 to S209 specifically specify the determination process of the target focused image. This step specifically identifies, by the preset character image feature, the determined person included in the current live image frame. Generally, when a video call is made, one or more subjects are often present in the current live image frame captured by the camera, so that the number of the selected persons actually included in the current live image frame can be identified according to the image characteristics of the person, and the recognition is performed. After the subject is present, the current pixel information of each subject in the current live image frame can also be determined, and the current pixel information can be specifically understood as a range of pixel values of all the pixels constituting one subject.

S204. Determine whether the subject is present in the acquired previous live image frame. If yes, go to step S205; if no, go to step S209.

This step can be used to determine whether the subject in the current live image frame also appears in the previous live image frame. Generally, different captured characters themselves have characteristics that are different from other captured characters (such as the clothes of the captured person). Color and wearing ornaments, etc., so it is possible to determine whether the subject is present in the previous real image frame according to the characteristics of the subject in the current live image frame, and the determined scene is not present in the current live image frame. When the person is photographed, the operation of step S209 can be performed; if there is the determined subject person, the operation of step S205 can be performed.

S205. Determine historical pixel information constituting the subject in the previous real image frame.

After determining that the determined subject is present in the previous real image frame, the step of determining the pixel position of the subject in the previous real image frame may be determined as the pixel position of the subject. Historical pixel information.

S206. Determine whether the current pixel information matches the historical pixel information. If not, perform the step. S207; if yes, step S208 is performed.

It should be noted that when the location of the smart terminal for the video call is fixed, the capture space corresponding to the camera of the smart terminal does not change, and the step may determine the historical pixel information of the determined person. Matches the current pixel information.

In this embodiment, if the subject is in the active dynamics, the historical pixel information in the previous real image frame and the current pixel information in the current live image frame cannot be completely matched, and the operation of step S207 can be performed. If the subject is in a stationary state, and its historical pixel information has a possibility of matching with the current pixel information, the operation of step S208 can be performed at this time.

S207. It is determined that the position of the subject is changed, and the subject is determined as the target focused image, and then step S210 is performed.

In this embodiment, when the historical pixel information of the object does not match the current pixel information, it may be determined that there is a change in the subject, and the subject may be determined as the target focused image, and the target may be determined. After the image is focused, the operation of step S210 is performed.

It should be noted that if there are a plurality of subjects whose positions have changed in the current live image frame, the subject with the lowest degree of matching of the historical pixel information and the current pixel information may be selected as the target focused image. For example, the degree of matching between the historical pixel information and the current pixel information may be specifically determined according to the number of matching pixel points, and the smaller the number of matched pixel points, the lower the matching degree.

S208. Determine average pixel information according to current pixel information of each subject, and determine an area corresponding to the average pixel information as the target focused image, and then perform step S210.

In this embodiment, if the current pixel information of each of the captured persons in the current live image frame matches the historical pixel information, it may be determined that the subject is still, and the step may be based on current pixel information of each of the captured characters. Determining the average pixel information of all the subjects in the current live image frame, thereby The area corresponding to the average pixel information is determined as the target focused image, and the operation of step S210 can be performed after the target focused image is determined.

S209. The preset focus pixel information is acquired, and the corresponding region of the focused pixel information in the current live image frame is determined as the target focused image, and then step S210 is performed.

This step processes the case where there is no object in the previous real image frame. The case where there is no object is generally that the captured current live image frame is the captured first frame, and there is no previous real scene. Image frame; or, the captured previous live image frame does not actually have a subject.

In this embodiment, when the situation in which the subject is not present is met, the preset focus pixel information may be acquired, and then the area corresponding to the focused pixel information is determined in the current live image frame, and the determined The area is focused on the image as the target, and the operation of step S210 can be performed after the target focused image is determined.

It should be noted that the captureable range of the camera disposed on the smart terminal is generally fixed, so that the present embodiment can set the focused pixel information according to the pixel information corresponding to the focused image determined during the capture of the historical image frame. .

S210. Determine plane coordinate information of the target focused image according to current pixel information of the target focused image in the current live image frame.

In this embodiment, the current live image frame is synthesized by the current image frame captured by at least two cameras, and the current live image frame includes spatial information of each image (plane coordinate information displayed on the screen and depth of rendering stereoscopic effect) value).

The embodiment may determine the corresponding plane coordinate information according to the current pixel information of the target focused image. Specifically, an average pixel coordinate value may be determined according to the pixel coordinate value of each pixel point in the current pixel information, which may be averaged in this embodiment. The pixel coordinate value is regarded as the plane coordinate information of the target focused image.

S211. Determine a depth value of the target focused image according to the depth information corresponding to the current pixel information.

Illustratively, according to the correspondence table of pixel points and depth information stored in advance, the present embodiment can determine the depth information corresponding to the average pixel coordinate value, and use the depth information as the depth value of the target focused image.

S212. Determine an actual focus distance of the target focused image to the camera according to the plane coordinate information and the depth value.

In this embodiment, the projection point of the target focused image in the stereoscopic space may be determined according to the plane coordinate information and the depth value, specifically, the pixel origin of the upper left corner of the screen of the smart terminal, and the target focused image is determined in the stereoscopic space. After the projection point in the middle, the actual distance value of the projection point to the pixel origin can be determined according to the plane coordinate information and the depth value, and the calculated actual distance value can be regarded as the actual focus distance of the target focused image to the camera.

S213. Determine a depth of field limit value of the current live image frame according to the actual focus distance and the acquired camera attribute parameter.

In this embodiment, the camera attribute parameters may include: a hyper focus distance and a lens focal length, wherein the hyper focus distance and the lens focal length are both determined by the type of camera used. Specifically, according to the actual focus distance and the acquired camera attribute parameters, and the calculation formula of the depth of field near the boundary

And the formula for calculating the far depth of the depth of field

To determine the current real image frame depth near the limit value and King reaching the limit value, wherein, S _near represents depth near the limit value, S _away showing scene reaching the limit value, H denotes hyperfocal camera distance, D denote the actual focusing distance, F represents the lens focal length of the camera.

Illustratively, when the camera property parameter is: f8, the hyperfocal distance is 6.25 meters (the blurring circle standard) When the lens focal length of the camera is 50 mm, if the real-focus distance is 4 meters, it can be determined according to the above formula that the depth of field has a near-limit value of 2.45 meters, and it is also possible to determine the depth of field limit. The value is 11.36 meters.

S214. Acquire image parameter information of an image region corresponding to a depth of field limit value, where the image parameter information includes: an image RGB ratio, a color contrast, and an image sharpness.

In this embodiment, the depth of field limit value is equivalent to the farthest distance that the camera can capture the image, and corresponds to the farthest image area in the current real image frame. The embodiment may determine the far depth limit value according to the depth of field. The image area and image parameter information of the image area, such as image RGB ratio, color contrast, and image sharpness.

In this embodiment, the image RGB ratio can be used to determine the brightness value of the image region; the color contrast can be a measure of different brightness levels between the brightest white and the darkest black in the light and dark regions of the image region. The larger the difference range is, the larger the color contrast is, and the smaller the difference range is, the smaller the color contrast is. The image sharpness can be understood as an index reflecting the image plane sharpness and the image edge sharpness, and the image sharpness is higher. The detail contrast on the image plane is also higher and looks more clear.

S215. Control, when the image parameter information does not meet the set standard parameter information, adjust image brightness, color contrast, and/or image sharpness of the image area, so that the image parameter information conforms to the standard parameter information.

In this embodiment, the image parameter information may be compared with the set standard parameter information, and the image brightness, the color contrast, and/or the image sharpness are respectively adjusted according to the comparison result, and finally the image parameters are finally obtained. The information conforms to the standard parameter information.

It can be understood that if the image area corresponding to the far depth limit value is a window image with higher brightness, after the image parameter information is adjusted, the display brightness of the window image can be appropriately reduced, thereby achieving clearness in the current real image frame. The purpose of displaying video participant image information.

The method for image processing provided by the second embodiment of the present invention embodies the process of acquiring the image frame, and at the same time, the process of determining the target focused image is embodied, and the process of determining the depth threshold of the depth of field and the far boundary of the depth of field are also embodied. The mediation process of the image area corresponding to the value. The method can acquire the image frame synthesized by the dual camera capture, and can determine the depth of field limit value of the image frame according to the depth information of the synthesized image frame and the determined target focus image, thereby performing an image on the region corresponding to the far depth limit value of the depth of field Mediation processing. By using the method, the determination and processing of the target area to be processed are efficiently realized, the overall processing of the entire image frame is effectively avoided, the flexibility of image processing is better, and the image processing efficiency during video call is improved. This further enhances the display effect of video participants on the smart terminal.

On the basis of the foregoing embodiment, after determining the subject of the current live image frame according to the character image feature, the embodiment further optimizes: performing brightness enhancement processing on the subject.

It should be noted that, based on the image processing described above in the embodiment, the adjustment processing of the image region corresponding to the depth of field limit value can be realized, so that the current live image frame has a clear image of the video participant. In addition, since the recognized person recognized in the current live image frame can be regarded as a video participant, the selected image area can be processed directly, and the recognized subject can be directly subjected to brightness enhancement processing.

Specifically, the specific pixel to be processed may be determined according to the current pixel information of the captured person and the corresponding depth information, and then the image parameter information of the to-be-processed area is determined, and the image parameter information is adjusted to improve the brightness of the captured person. Improved, so that the subject can have a better display in the current live image frame.

Embodiment 3

FIG. 3 is a structural block diagram of an apparatus for image processing according to Embodiment 3 of the present invention. The device is suitable for image processing of captured image frames during a video call, wherein the device can be implemented by software and/or hardware and is generally integrated on a smart terminal having a video call function. As shown in FIG. 3, the apparatus includes: a live image acquisition module 31, a focused image determination module 32, a depth of field limit determination module 33, and an image parameter adjustment module 34.

The real image acquisition module 31 is configured to acquire a current real image frame captured by the camera;

a focused image determining module 32, configured to determine a target focused image in the current live image frame;

a depth of field limit determining module 33, configured to determine a depth of field limit value of the current live image frame according to the target focused image;

The image parameter adjustment module 34 is configured to perform image parameter information adjustment processing on the image region corresponding to the depth of field limit value.

In the present embodiment, the device first acquires the current live image frame captured by the camera through the live image acquisition module 31; then determines the target focused image in the current live image frame by the focused image determination module 32; and then passes the depth of field determination module 33 Determining the depth of field limit value of the current live image frame according to the target focused image; finally, the image parameter adjustment module 34 performs the adjustment processing of the image parameter information on the image region corresponding to the far depth limit value.

An apparatus for image processing according to Embodiment 3 of the present invention can perform adjustment processing on a partial image in an image frame captured during a video call, and efficiently realizes determination and processing of a target area to be processed, and further increases an image. The flexibility of processing effectively improves the display effect of video participants on smart terminals.

Further, the real-time image obtaining module 31 is specifically configured to: acquire at least two photos by using The current image frame captured by the image headers is respectively subjected to image synthesis processing on the at least two current image frames respectively captured to obtain a current live image frame; wherein each pixel point in the current live image frame has corresponding depth of field information.

On the basis of the above optimization, the focused image determining module 32 includes:

a subject determining unit, configured to determine a subject of the current live image frame according to a character image feature, and determine current pixel information constituting the subject; an information determining unit, configured to acquire the previous real scene Determining whether the subject is present in the image frame; a first execution unit, configured to determine historical pixel information constituting the subject in the previous live image frame when the subject is present, and Determining whether the current pixel information matches the historical pixel information, and if not, determining that the position of the subject is changed, determining the subject to be a target focused image; if so, according to each of the subjects The current pixel information determines the average pixel information, and the region corresponding to the average pixel information is determined as the target focused image; and the second execution unit is configured to acquire the preset focused pixel information when the captured person does not exist, The focused pixel information is determined as a target focused image in a corresponding region in the current live image frame.

Further, the focused image determining module 32 further includes: a subject processing unit, configured to perform brightness enhancement processing on the subject after the subject of the current live image frame is determined according to the character image feature .

On the basis of the foregoing embodiment, the depth of field determination module 33 is configured to: determine plane coordinate information of the target focused image according to current pixel information of the target focused image in the current live image frame; Determining, according to the depth of field information corresponding to the current pixel information, a depth value of the target focused image; determining an actual focus distance of the target focused image to the camera according to the plane coordinate information and the depth value; The actual focus distance and the acquired camera properties a parameter that determines a depth of field limit value of the current live image frame.

Further, the image parameter adjustment module 34 is specifically configured to: acquire image parameter information of an image region corresponding to the depth of field limit value, where the image parameter information includes: an image RGB ratio, a color contrast, and an image sharpness; When the image parameter information does not conform to the set standard parameter information, controlling to adjust image brightness, color contrast, and/or image sharpness of the image area, so that the image parameter information conforms to the standard parameter information.

Embodiment 4

The fourth embodiment of the present invention further provides an intelligent conference terminal, including: at least two cameras with parallel optical axes, and an apparatus for image processing provided by the foregoing embodiments of the present invention. The image processing can be performed by the image processing methods provided in the first embodiment and the second embodiment.

In this embodiment, the smart conference terminal belongs to a type of electronic device having a video call function, wherein the smart conference terminal is integrated with a video call system, and at the same time, at least two cameras with parallel optical axes and the above The apparatus for image processing provided by the embodiment.

After the device for image processing provided by the foregoing embodiment of the present invention is integrated in the smart conference terminal, when a video call is made with other smart terminals having a video call function, the partial image in the current live view image frame captured in real time can be performed. The adjustment and processing of the image parameter information effectively improves the display effect of the video participant on the intelligent conference terminal, and further improves the user experience of the intelligent conference terminal.

One of ordinary skill in the art can understand that all or part of the steps of implementing the above embodiments may be completed by a program to instruct related hardware, and the program may be stored in a computer readable storage medium, and the program is executed. The following steps are included: obtaining the current captured by the camera Realizing an image frame, and determining a target focused image in the current real image frame; determining a depth of field limit value of the current live image frame according to the target focused image; and performing an image on the image region corresponding to the depth of field limit value Adjustment processing of parameter information. The storage medium is, for example, a ROM/RAM, a magnetic disk, an optical disk, or the like.

Note that the above are only the preferred embodiments of the present invention and the technical principles applied thereto. Those skilled in the art will appreciate that the present invention is not limited to the specific embodiments described herein, and that various modifications, changes and substitutions may be made without departing from the scope of the invention. Therefore, the present invention has been described in detail by the above embodiments, but the present invention is not limited to the above embodiments, and other equivalent embodiments may be included without departing from the inventive concept. The scope is determined by the scope of the appended claims.

Claims

A method of image processing, comprising:

Obtaining a current live image frame captured by the camera and determining a target focused image in the current live image frame;

Determining a depth of field limit value of the current live image frame according to the target focused image;

The image region information is subjected to adjustment processing on the image region corresponding to the depth of field limit value.
The method according to claim 1, wherein the acquiring a current live image frame captured by the camera comprises:

Obtaining a current image frame captured by at least two cameras respectively;

Performing image synthesis processing on at least two current image frames respectively captured to obtain a current live image frame;

Wherein, each pixel in the current live image frame has corresponding depth of field information.
The method according to claim 2, wherein said determining a target focused image in said current live image frame comprises:

Determining a subject in the current live image frame according to a character image feature, and determining current pixel information constituting the subject;

Determining whether the subject is present in the acquired previous live image frame;

If the subject is present, determining historical pixel information constituting the subject in the previous live image frame, and determining whether the current pixel information matches the historical pixel information, and if not, Determining that the position of the subject changes, determining the subject as a target focused image; if yes, determining average pixel information according to current pixel information of each of the subjects, determining an area corresponding to the average pixel information Focus the image for the target;

If the subject is not present, acquiring preset focus pixel information, the focused pixel The corresponding area of the information in the current live image frame is determined as the target focused image.
The method according to claim 3, further comprising: after determining the subject of the current live image frame according to the character image feature, further comprising:

The subject is subjected to brightness enhancement processing.
The method according to claim 2, wherein the determining a depth of field limit value of the current live image frame according to the target focused image comprises:

Determining plane coordinate information of the target focused image according to current pixel information of the target focused image in the current live image frame;

Determining a depth value of the target focused image according to the depth information corresponding to the current pixel information;

Determining an actual focus distance of the target focused image to the camera according to the plane coordinate information and the depth value;

Determining a depth of field limit value of the current live image frame according to the actual focus distance and the acquired camera attribute parameter.
The method according to claim 1, wherein the performing image parameter information adjustment processing on the image region corresponding to the depth of field limit value includes:

And acquiring image parameter information of the image region corresponding to the depth of field limit value, where the image parameter information includes: image RGB proportion, color contrast, and image sharpness;

When the image parameter information does not conform to the set standard parameter information, controlling to adjust image brightness, color contrast, and/or image sharpness of the image area, so that the image parameter information conforms to the standard parameter information.
An apparatus for image processing, comprising:

a real image acquisition module, configured to acquire a current live image frame captured by the camera;

a focused image determining module, configured to determine a target focused image in the current live image frame;

a depth of field limit determining module, configured to determine a depth of field limit value of the current live image frame according to the target focused image;

The image parameter adjustment module is configured to perform image parameter information adjustment processing on the image region corresponding to the depth of field limit value.
The device according to claim 7, wherein the real-time image acquisition module is specifically configured to:

Obtaining a current image frame captured by at least two cameras;

Performing image synthesis processing on at least two current image frames respectively captured to obtain a current live image frame;

Wherein, each pixel in the current live image frame has corresponding depth of field information.
The apparatus according to claim 8, wherein the focused image determining module comprises:

a subject determining unit, configured to determine a subject in the current live image frame according to the character image feature, and determine current pixel information constituting the subject;

An information determining unit, configured to determine whether the subject is present in the acquired previous live image frame;

a first execution unit, configured to determine historical pixel information constituting the subject in the previous live image frame when the subject is present, and determine whether the current pixel information is related to the historical pixel Information matching, if not, determining that the position of the subject changes, determining the subject as a target focused image; if yes, determining average pixel information according to current pixel information of each subject, The area corresponding to the average pixel information is determined as the target focused image;

a second execution unit, configured to acquire preset focus pixel information when the subject is not present, and determine a corresponding area of the focused pixel information in the current live image frame as a target focused image.
An intelligent conference terminal comprising: at least two cameras parallel to an optical axis, characterized by further comprising: an apparatus for image processing according to any one of claims 7-9.