WO2019042243A1 - 一种图像遮蔽方法、装置、设备及系统 - Google Patents

一种图像遮蔽方法、装置、设备及系统 Download PDF

Info

Publication number
WO2019042243A1
WO2019042243A1 PCT/CN2018/102467 CN2018102467W WO2019042243A1 WO 2019042243 A1 WO2019042243 A1 WO 2019042243A1 CN 2018102467 W CN2018102467 W CN 2018102467W WO 2019042243 A1 WO2019042243 A1 WO 2019042243A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
preset
masked
value
target area
Prior art date
Application number
PCT/CN2018/102467
Other languages
English (en)
French (fr)
Inventor
陈益伟
Original Assignee
杭州海康威视数字技术股份有限公司
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 杭州海康威视数字技术股份有限公司 filed Critical 杭州海康威视数字技术股份有限公司
Publication of WO2019042243A1 publication Critical patent/WO2019042243A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/002D [Two Dimensional] image generation
    • G06T11/40Filling a planar surface by adding surface attributes, e.g. colour or texture
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T9/00Image coding

Definitions

  • the present application relates to the field of image processing technologies, and in particular, to an image masking method, apparatus, device, and system.
  • This sensitive area is typically an image area that involves privacy or requires privacy.
  • This sensitive area is usually an image area that involves privacy or requires privacy.
  • it is usually necessary to mosaic or blacken the face; or, for some information in the image that needs to be kept secret, such as license plates and ID cards, it is usually required to be mosaic or blackened.
  • the image occlusion scheme includes: the terminal acquires a video stream collected by the collection device; and when the terminal displays the video image in the video stream, setting a shadow image in a region where a face or other secret information may appear in the image, the mask image may be Mosaic images or black images so you can mask sensitive areas of the video image.
  • the position of the masked picture in the video image is usually preset, and if the sensitive area does not fall at the position where the masked picture is located, effective masking cannot be formed. It can be seen that the above scheme is applied, and the shading effect is poor.
  • An object of the embodiments of the present application is to provide an image masking method, device, device and system to improve the shielding effect.
  • an image masking method including:
  • the image data in the target area is modified into preset mask data to obtain a masked image.
  • the detecting the target area to be masked in the image may include:
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content;
  • a target area to be masked in the image is determined.
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, including:
  • the RGB value of each pixel in the target area to a first set of preset values to obtain a masked image, where the first set of preset values includes a preset R value, a preset G value, and a preset Set the B value;
  • the image data in the target area is modified into the preset masking data to obtain the masked image, which may include:
  • a YUV value of each pixel in the target area to a second set of preset values to obtain a masked image, the second set of preset values including a preset Y value, a preset U value, and a preset V value.
  • the second set of preset values is determined by the following steps:
  • the set of RGB preset values is converted into a second set of preset values by using a conversion relationship between the RGB format data and the YUV format data.
  • the method may further include:
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • the method may further include:
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • an embodiment of the present application further provides an image screening apparatus, including:
  • a detecting module configured to detect, in each frame image, a target area to be shielded in the image
  • a modifying module configured to modify image data in the target area to preset masking data to obtain a masked image.
  • the detecting module is specifically configured to:
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content; determining a target area to be masked in the image based on the output result.
  • the modifying module is specifically configured to:
  • the RGB value of each pixel in the target area is set to a first group preset value to obtain a masked image, where the first set of preset values is Contains preset R value, preset G value, preset B value;
  • the YUV value of each pixel in the target area is set to a second set of preset values to obtain a masked image, and the second set of preset values includes Preset Y value, preset U value, preset V value.
  • the device may further include:
  • a determining module for determining a set of RGB preset values; converting the set of RGB preset values into a second set of preset values by using a conversion relationship between the RGB format data and the YUV format data.
  • the device further includes: a display module, or further comprising: a packaging module and a sending module, where
  • a display module configured to display the shaded image
  • a packaging module configured to encode and encapsulate the masked image by using a preset encoding encapsulation format to obtain a packaged image
  • a sending module configured to send the encapsulated image to a player for playing.
  • the device may further include:
  • a display module configured to display the shaded image
  • a judging module configured to judge whether the shading effect of the displayed image is qualified; if yes, triggering the encapsulation module;
  • a packaging module configured to encode and encapsulate the masked image by using a preset encoding encapsulation format to obtain a packaged image
  • a sending module configured to send the encapsulated image to a player for playing.
  • an embodiment of the present application further provides an electronic device, including a processor and a memory, where the memory is used to store executable program code, and the processor operates by reading executable program code stored in the memory.
  • a program corresponding to the program code is executable to perform any of the image masking methods described above.
  • an embodiment of the present application further provides a computer readable storage medium, where the computer readable storage medium stores a computer program, and when the computer program is executed by a processor, implementing any of the image blocking methods described above. .
  • an embodiment of the present application further discloses an executable program code for being executed to perform any of the image masking methods described above.
  • an image screening system including: an acquisition device and a processing device;
  • the collecting device is configured to collect a video stream, and send the collected video stream to the processing device;
  • the processing device is configured to receive a video stream sent by the collection device, and detect, for each frame image in the received video stream, a target area to be masked in the image; and display an image in the target area
  • the data is modified to preset mask data, and the shaded image is obtained.
  • the system further includes: a playback device,
  • the processing device is further configured to: encode and encapsulate the masked image by using a preset encoding encapsulation format, to obtain a encapsulated image; and send the encapsulated image to the playback device;
  • the playback device is configured to receive the encapsulated image sent by the processing device, decapsulate the encapsulated image, obtain a masked image, and play the masked image.
  • detecting a target area to be masked in each frame image, modifying image data in the target area to preset masking data, and obtaining a masked image, which is visible in the present solution is not based on
  • the fixed position masking image masks the sensitive area in the image, but first detects the position of the sensitive area (target area) in the image, and then modifies the image data at the position, even if the sensitive area moves, it can still detect The sensitive area is removed and shielded to improve the shadowing effect.
  • FIG. 1 is a schematic flowchart of an image masking method according to an embodiment of the present application
  • FIG. 2 is a schematic diagram of an image coordinate system provided by an embodiment of the present application.
  • FIG. 3 is a schematic diagram of an application scenario provided by an embodiment of the present application.
  • FIG. 4 is a schematic structural diagram of an image shielding apparatus according to an embodiment of the present disclosure.
  • FIG. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure.
  • FIG. 6 is a schematic diagram of a first structure of an image screening system according to an embodiment of the present application.
  • FIG. 7 is a schematic diagram of a second structure of an image screening system according to an embodiment of the present application.
  • FIG. 8 is a schematic diagram of a third structure of an image screening system according to an embodiment of the present application.
  • an embodiment of the present application provides an image masking method, apparatus, device, and system.
  • the method can be applied to a device having an image processing function, which is not limited.
  • FIG. 1 is a schematic flowchart of an image masking method according to an embodiment of the present disclosure, including:
  • S102 Modify image data in the target area to preset mask data to obtain a masked image.
  • FIG. 1 The following is a detailed description of the embodiment shown in FIG. 1:
  • the device (or the execution body, hereinafter referred to as the electronic device) that implements the solution may be an image acquisition device or other electronic device that is in communication with the image acquisition device. If the electronic device is an image capture device, the image capture device may apply the scheme for image masking for each frame image in the collected video stream. If the electronic device is another electronic device communicatively coupled to the image capture device, the electronic device can obtain the captured video stream from the image capture device, and apply the present solution to the image for each frame image in the video stream. Shaded.
  • the single image may be masked, which is not limited.
  • S101 may include:
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content; determining a target area to be masked in the image based on the output result.
  • image content to be masked is consistent with the content in the target area to be masked in S101.
  • the content in the target area to be masked in S101 is also a human face;
  • the image content to be masked is a license plate, the target to be masked in S101
  • the content in the area is also the license plate.
  • the face image sample is learned in advance using the depth learning algorithm to obtain a face detection model; in S101, the image is input to the face detection model, and the image can be determined according to the output result.
  • the face area in the image is learned in advance using the depth learning algorithm to obtain a face detection model; in S101, the image is input to the face detection model, and the image can be determined according to the output result. The face area in the image.
  • the depth learning algorithm is used in advance to learn the license plate image sample to obtain a license plate detection model; in S101, the image is input to the license plate. The model is detected and the license plate area in the image can be determined based on the output.
  • the target area to be masked in the image can be detected by other means:
  • a face feature point model may be established in advance. Specifically, the feature points of the nose, the eyes, the mouth, and the like can be extracted, and the face feature point model is constructed according to the face proportion. The image is matched with the face feature point model, and the area that is successfully matched is the face area, that is, the target area to be masked.
  • the license plate recognition technology can be used to identify the license plate area in the image, and the like, and the specific detection manner is not limited.
  • a target rectangular frame may be obtained, and the image area in the target rectangular frame is the target area.
  • the coordinate system can be constructed with a corner point of the image as the origin.
  • the coordinate system is constructed with the corner point of the upper left corner of the image as the origin, and the coordinates of the four corner points of the target rectangular frame are determined as :A(W/3, H/3), B(2W/3, H/3), C(W/3, 2H/3), D(2W/3, 2H/3), the coordinates are at W/
  • the pixel points in the 3-2W/3, H/3-2H/3 section are used as the pixel points in the target area for subsequent processing.
  • S102 Modify image data in the target area to preset mask data to obtain a masked image.
  • the image has multiple formats, such as RGB (Red Green Blue) format, or YUV (Brightness and Color Difference) format.
  • RGB Red Green Blue
  • YUV YUV
  • the RGB value of each pixel in the target area may be set to a first set of preset values to obtain a masked image, where the first set of preset values includes Preset R value, preset G value, preset B value.
  • the target area is an area with the same color, so that the original picture content of the target area is obscured, and the shaded image is obtained.
  • resetting the RGB values of each pixel in the target area to the same value does not mean that the R value, the G value, and the B value of one pixel point are the same, but rather, The R values of each pixel are the same, the G values are the same, and the B values are the same. For example, the R value of each pixel in the target area is adjusted to 100, the G value is adjusted to 120, and the B value is adjusted to 130, and the original picture content of the target area is masked.
  • the RGB values of each pixel in the target area may be all set to 0, or all set to 255, or all set to a certain value, such that the R value, the G value, and the B value of one pixel point are the same.
  • the R value of each pixel is the same, the G values are the same, and the B values are the same, and the original picture content can also be masked.
  • the first set of preset values includes a preset R value, a preset G value, and a preset B value
  • the three values may be the same or different, and the three values may be any values of 0-255; Adjusting the R value of each pixel in the target area to the preset R value in the first group preset value, and adjusting the G value of each pixel in the target area to the first group preset value.
  • the preset G value adjusts the B value of each pixel in the target area to the preset B value in the first group preset value, so that the effect of masking the image content can be achieved.
  • a set of numerical intervals may be set, where the numerical interval includes a preset R value interval, a preset G value interval, and a preset B value interval; and the R value of each pixel in the target region is adjusted to the The R value interval is preset, and the G value of each pixel in the target area is adjusted to the preset G value interval, and the B value of each pixel in the target area is adjusted to the preset B value interval.
  • the RGB values of each pixel in the target area are adjusted to the same set of numerical values, that is, the RGB values of each pixel are adjusted to approximate values, so that the target area is a similar color area.
  • the original picture content of the target area is obscured, and the masked image is obtained.
  • the YUV value of each pixel in the target area may be set to a second set of preset values to obtain a masked image, and the second set of preset values includes a pre- Set Y value, preset U value, preset V value.
  • the second set of preset values may be preset according to the experience value.
  • the second set of preset values is directly read, and the preset Y value, the preset U value, and the preset V value are obtained, and the target area is obtained.
  • the Y value of each pixel in the middle is set to the preset Y value
  • the U value of each pixel in the target area is set to the preset U value
  • the V value of each pixel in the target area is set. For this preset V value.
  • the second set of preset values may be set in other manners.
  • the YUV value of the image may be repeatedly adjusted.
  • the adjusted YUV value is recorded as the second. Group preset value.
  • the second set of preset values may also be determined by the following steps:
  • Determining a set of RGB preset values converting the set of RGB preset values into a second set of preset values according to a conversion relationship between the RGB format data and the YUV format data.
  • the value of YUV can be calculated, and the obtained value of YUV is taken as the second set of preset values.
  • the preset occlusion data in S102 may also be mosaic data, that is, the image data in the target area is replaced with mosaic data.
  • the above scheme can be applied to each frame of the video stream: the target area is automatically detected, and the image data of the target area is automatically modified, so that even in the process of processing a video, even if the sensitive area moves, the sensitivity can be detected.
  • the area is masked and the shading effect is improved.
  • the sensitive area in the image is not masked based on the fixed position, but the position of the sensitive area (target area) in the image is detected first. Then, the image data at the position is modified, and even if the sensitive area moves, the sensitive area can be detected and masked, thereby improving the shadowing effect; in the second aspect, if only the shadow image is set during the image display process
  • the image in the transmission process is still an unmasked image, and the privacy content in the image may still be leaked, and the security is low, and the image data itself is modified by the scheme, and the image in the transmission process is the shadowed image, and the image is improved. The security of the data.
  • the occluded image may be displayed, or the occluded image may be encoded and encapsulated by using a preset coding encapsulation format to obtain a packaged image;
  • the encapsulated image is sent to the player for playback.
  • the electronic device itself may display the masked image, or may encapsulate the masked image and send it to the player for display.
  • the modified YUV format data can be encoded into H.264 format after the YUV value of each pixel in the target area is modified to the second set of preset values. Then, the encoded data is encapsulated into PS (Program Stream) data, and the PS data is sent to the player for playback.
  • PS Program Stream
  • the electronic device first displays the masked image to determine whether the masking effect of the displayed image is qualified; if yes, encoding the masked image by using a preset encoding encapsulation format. Encapsulating, obtaining a packaged image; sending the encapsulated image to a player for playback.
  • the electronic device may display the masked image obtained in S102 to the user. If the user determines that the masking effect of the displayed image is acceptable, the first instruction is sent to the electronic device, and the first instruction may be a confirmation instruction and play. Instructions, etc. After receiving the first instruction, the electronic device encodes and encapsulates the masked image obtained in S102 by using a preset encoding encapsulation format to obtain a packaged image; and sends the encapsulated image to the player for playing.
  • the second instruction may be sent to the electronic device, and the second instruction may be a stop instruction or the like. After receiving the second instruction, the electronic device may stop masking other images in the video stream.
  • the electronic device displays the masked image obtained in S102, and the electronic device can further judge the displayed image to determine whether there is a sensitive area in the displayed image that needs to be shielded. If not, then Indicates that the masking effect of the displayed image is acceptable.
  • the electronic device encodes and encapsulates the masked image obtained in S102 by using a preset encoding and encapsulating format to obtain a packaged image; and sends the encapsulated image to the player for playback.
  • Further determining whether there is a sensitive area in the displayed image that needs to be masked may be detected by using the above detection model, or if the sensitive area is a human face, the face feature point model may be used for matching, if detected in the masked image In the face area, it is determined that the shading effect is unsatisfactory.
  • the sensitive area is a license plate
  • the license plate recognition technology may be used to determine whether there is a license plate number in the image, and if the license plate area is detected in the shaded image, it is determined that the shielding effect is unqualified.
  • the collection device collects the video stream in real time; the collection device sends the video stream to the processing device through the switching device; or the collection device can directly send the video stream to the processing device.
  • the processing device receives the video stream and decodes the video stream to obtain YUV data.
  • the processing device inputs the image into the detection model obtained in advance for each frame image in the YUV data, and outputs a target rectangular frame, and the image region in the target rectangular frame is the target region to be masked.
  • the processing device sets the YUV value of each pixel in the target area to the second set of preset values.
  • the process of predetermining the second set of preset values may include: determining a set of RGB preset values; using the conversion relationship between the RGB format data and the YUV format data, inversely solving the YUV value of the pixel, and the solutiond value is the second Group preset value.
  • the processing device displays the image processed in step 4 as the masked image. If the masking effect of the displayed image is acceptable, the processing device encodes and encapsulates the masked image by using a preset encoding encapsulation format, and obtains the encapsulated image. Image.
  • the processing device sends the encapsulated image to the player for playback.
  • the player can be a built-in player in the processing device, or can be other playback devices, and is not limited.
  • the sensitive area in the image is not masked based on the fixed position, but the position of the sensitive area (target area) in the image is detected first, and then the position is The image data is modified, and even if the sensitive area moves, the sensitive area can be detected and masked, thereby improving the shadowing effect;
  • the second aspect if only the masking picture is set during the image display, the image during transmission is Still unmasked image, it is still possible to leak the privacy content in the image, and the security is low, and the scheme modifies the image data itself, and the image in the transmission process is the masked image, which improves the security of the data;
  • the processing device masks the image, and the processing device reduces the cost of the collecting device compared to the processing device.
  • the embodiment of the present application further provides an image screening device.
  • FIG. 4 is a schematic structural diagram of an image screening apparatus according to an embodiment of the present disclosure, including:
  • a detecting module 401 configured to detect, in each frame image, a target area to be masked in the image
  • the modifying module 402 is configured to modify image data in the target area to preset masking data to obtain a masked image.
  • the detecting module 401 can be specifically configured to:
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content; determining a target area to be masked in the image based on the output result.
  • the modifying module 402 may be specifically configured to:
  • the RGB value of each pixel in the target area is set to a first group preset value to obtain a masked image, where the first set of preset values is Contains preset R value, preset G value, preset B value;
  • the YUV value of each pixel in the target area is set to a second set of preset values to obtain a masked image, and the second set of preset values includes Preset Y value, preset U value, preset V value.
  • the device may further include:
  • a determining module (not shown) for determining a set of RGB preset values; converting the set of RGB preset values into a second set of preset values by using a conversion relationship between the RGB format data and the YUV format data.
  • the device may further include: a display module (not shown), or further comprising: a packaging module and a sending module (not shown), wherein
  • a display module configured to display the shaded image
  • a packaging module configured to encode and encapsulate the masked image by using a preset encoding encapsulation format to obtain a packaged image
  • a sending module configured to send the encapsulated image to a player for playing.
  • the device may further include: a display module, a determination module, a packaging module, and a sending module (not shown), where
  • a display module configured to display the shaded image
  • a judging module configured to judge whether the shading effect of the displayed image is qualified; if yes, triggering the encapsulation module;
  • a packaging module configured to encode and encapsulate the masked image by using a preset encoding encapsulation format to obtain a packaged image
  • a sending module configured to send the encapsulated image to a player for playing.
  • the position of the sensitive area (target area) in the image is detected first, and then the position is detected.
  • the image data at the location is modified, and even if the sensitive area moves, the sensitive area can be detected and masked, thereby improving the shadowing effect;
  • the masking picture is set only during the image display, the transmission process The image is still unmasked, and it is still possible to leak the privacy content in the image, and the security is low.
  • the scheme modifies the image data itself, and the image in the transmission process is the masked image, which improves the security of the data.
  • the processing device masks the image, and the processing device reduces the cost of the collecting device compared to the processing device.
  • FIG. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application, including: a processor 501 and a memory 502, where the memory 502 is configured to store executable program code;
  • the processor 501 runs a program corresponding to the executable program code by reading the executable program code stored in the memory 502 for performing the following steps:
  • the image data in the target area is modified into preset mask data to obtain a masked image.
  • the processor 501 is further configured to perform the following steps:
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content;
  • a target area to be masked in the image is determined.
  • the processor 501 is further configured to perform the following steps:
  • the RGB value of each pixel in the target area is set to a first group preset value to obtain a masked image, where the first set of preset values is Contains preset R value, preset G value, preset B value;
  • the YUV value of each pixel in the target area is set to a second set of preset values to obtain a masked image, and the second set of preset values includes Preset Y value, preset U value, preset V value.
  • the processor 501 is further configured to perform the following steps:
  • the set of RGB preset values is converted into a second set of preset values by using a conversion relationship between the RGB format data and the YUV format data.
  • the processor 501 is further configured to perform the following steps:
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, the masked image is displayed;
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • the processor 501 is further configured to perform the following steps:
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, the masked image is displayed;
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • the memory mentioned in the above electronic device may include a random access memory (RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage.
  • NVM non-volatile memory
  • the memory may also be at least one storage device located away from the aforementioned processor.
  • the above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; or may be a digital signal processing (DSP), dedicated integration.
  • CPU central processing unit
  • NP network processor
  • DSP digital signal processing
  • ASIC Application Specific Integrated Circuit
  • FPGA Field-Programmable Gate Array
  • the embodiment of the present application further provides a computer readable storage medium, where the computer readable storage medium stores a computer program, and when the computer program is executed by the processor, the following steps are implemented:
  • the image data in the target area is modified into preset mask data to obtain a masked image.
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content;
  • a target area to be masked in the image is determined.
  • the RGB value of each pixel in the target area is set to a first group preset value to obtain a masked image, where the first set of preset values is Contains preset R value, preset G value, preset B value;
  • the YUV value of each pixel in the target area is set to a second set of preset values to obtain a masked image, and the second set of preset values includes Preset Y value, preset U value, preset V value.
  • the set of RGB preset values is converted into a second set of preset values by using a conversion relationship between the RGB format data and the YUV format data.
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, the masked image is displayed;
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, the masked image is displayed;
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • Embodiments of the present application also provide an executable program code for being executed to perform any of the image masking methods described above.
  • An embodiment of the present application further provides an image screening system, as shown in FIG. 6, comprising: an acquisition device and a processing device;
  • An acquiring device configured to collect a video stream, and send the collected video stream to the processing device;
  • a processing device configured to receive a video stream sent by the collection device, detect, for each frame image in the received video stream, a target area to be masked in the image; modify image data in the target area Obscured data for the preset, and the shaded image is obtained.
  • the system may further include a playback device, as shown in FIG. 7.
  • the processing device may further be configured to encode the masked image by using a preset encoding encapsulation format. Encapsulating, obtaining a packaged image; transmitting the encapsulated image to the playback device;
  • a playback device configured to receive the encapsulated image sent by the processing device; decapsulate the encapsulated image to obtain a masked image; and play the masked image.
  • the processing device may have a built-in player, and the processing device may encode and encapsulate the masked image by using a preset encoding encapsulation format to obtain a packaged image;
  • the encapsulated image is sent to the built-in player for playback; the player decapsulates the encapsulated image to obtain a masked image; and plays the masked image.
  • the processing device can also be used to:
  • the detection model is: using a depth learning algorithm to learn a preset image sample, where the preset image sample includes Masked image content; determining a target area to be masked in the image based on the output result.
  • the processing device can also be used to:
  • the RGB value of each pixel in the target area is set to a first group preset value to obtain a masked image, where the first set of preset values is Contains preset R value, preset G value, preset B value;
  • the YUV value of each pixel in the target area is set to a second set of preset values to obtain a masked image, and the second set of preset values includes Preset Y value, preset U value, preset V value.
  • the processing device can also be used to:
  • the set of RGB preset values is converted into a second set of preset values by using a conversion relationship between the RGB format data and the YUV format data.
  • the processing device can also be used to:
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, the masked image is displayed;
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • the processing device can also be used to:
  • the image data in the target area is modified into preset mask data, and the masked image is obtained, the masked image is displayed;
  • the masked image is encoded and encapsulated by using a preset encoding encapsulation format to obtain a packaged image; and the encapsulated image is sent to a player for playing.
  • the first aspect is not based on the fixed position of the fixed position to mask the sensitive area in the image, but first detecting the position of the sensitive area (target area) in the image, and then at the position The image data is modified, and even if the sensitive area moves, the sensitive area can be detected and masked, thereby improving the shadowing effect;
  • the second aspect if only the masking picture is set during the image display, the image during transmission is Still unmasked image, it is still possible to leak the privacy content in the image, and the security is low, and the scheme modifies the image data itself, and the image in the transmission process is the masked image, which improves the security of the data;
  • the processing device masks the image, and the processing device reduces the cost of the collecting device compared to the processing device.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
  • Image Processing (AREA)

Abstract

一种图像遮蔽方法、装置、设备及系统,该方法包括:针对每一帧图像,检测该图像中待遮蔽的目标区域(S101);将该目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像(S102)。第一方面,该方法并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生移动,仍能够检测出敏感区域,并对其遮蔽,提高了遮蔽效果;第二方面,如果只是在图像展示过程中设置遮蔽图片,传输过程中的图像仍为未遮蔽的图像,仍有可能泄露图像中的隐私内容,安全性较低,而该方法对图像数据本身进行修改,传输过程中的图像为遮蔽后的图像,提高了数据的安全性。

Description

一种图像遮蔽方法、装置、设备及系统
本申请要求于2017年8月29日提交中国专利局、申请号为201710754705.6、发明名称为“一种图像遮蔽方法、装置、设备及系统”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及图像处理技术领域,特别是涉及一种图像遮蔽方法、装置、设备及系统。
背景技术
在图像展示过程中,通常需要对图像中的敏感区域进行遮蔽。该敏感区域通常为涉及隐私或者需要保密的图像区域。比如,在播放视频时,为了保护当事人的隐私,通常需要将人脸打上马赛克或者涂黑;或者,对于图像中一些需要保密的信息,如车牌、身份证等也通常需要打上马赛克或者涂黑。
通常,图像遮蔽方案包括:终端获取采集设备采集的视频流;终端对视频流中的视频图像进行展示时,在图像中可能出现人脸或其他保密信息的区域设置遮蔽图片,该遮蔽图片可以为马赛克图片或者黑色图片,这样便可以遮蔽掉视频图像中的敏感区域。
但是,该遮蔽图片在视频图像中的位置通常是预先设定的,如果敏感区域没有落在该遮蔽图片所在的位置,则不能形成有效遮蔽。可见,应用上述方案,遮蔽效果差。
发明内容
本申请实施例的目的在于提供一种图像遮蔽方法、装置、设备及系统,以提高遮蔽效果。
为达到上述目的,本申请实施例提供了一种图像遮蔽方法,包括:
针对每一帧图像,检测所述图像中待遮蔽的目标区域;
将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
可选的,所述检测所述图像中待遮蔽的目标区域,可以包括:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;
根据所述输出结果,确定所述图像中待遮蔽的目标区域。
可选的,在所述图像数据为RGB格式的情况下,所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像,包括:
将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
在所述图像数据为YUV格式的情况下,所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像,可以包括:
将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
可选的,采用如下步骤确定所述第二组预设值:
确定一组RGB预设值;
利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
可选的,在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,还可以包括:
对所述遮蔽后的图像进行展示;
或者,
利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
可选的,在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,还可以包括:
对所述遮蔽后的图像进行展示;
判断所展示图像的遮蔽效果是否合格;
如果是,利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
为达到上述目的,本申请实施例还提供了一种图像遮蔽装置,包括:
检测模块,用于针对每一帧图像,检测所述图像中待遮蔽的目标区域;
修改模块,用于将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
可选的,所述检测模块,具体可以用于:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;根据所述输出结果,确定所述图像中待遮蔽的目标区域。
可选的,所述修改模块,具体可以用于:
在所述图像数据为RGB格式的情况下,将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
在所述图像数据为YUV格式的情况下,将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
可选的,所述装置还可以包括:
确定模块,用于确定一组RGB预设值;利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
可选的,所述装置还包括:展示模块,或者还包括:封装模块和发送模块,其中,
展示模块,用于对所述遮蔽后的图像进行展示;
封装模块,用于利用预设编码封装格式,对所述遮蔽后的图像进行编码 封装,得到封装后的图像;
发送模块,用于将所述封装后的图像发送至播放器进行播放。
可选的,所述装置还可以包括:
展示模块,用于对所述遮蔽后的图像进行展示;
判断模块,用于判断所展示图像的遮蔽效果是否合格;如果是,触发封装模块;
封装模块,用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;
发送模块,用于将所述封装后的图像发送至播放器进行播放。
为达到上述目的,本申请实施例还提供了一种电子设备,包括处理器和存储器,其中,存储器用于存储可执行程序代码,处理器通过读取存储器中存储的可执行程序代码来运行与可执行程序代码对应的程序,以用于执行上述任一种图像遮蔽方法。
为达到上述目的,本申请实施例还提供了一种计算机可读存储介质,所述计算机可读存储介质内存储有计算机程序,所述计算机程序被处理器执行时实现上述任一种图像遮蔽方法。
为达到上述目的,本申请实施例还公开了一种可执行程序代码,所述可执行程序代码用于被运行以执行上述任一种图像遮蔽方法。
为达到上述目的,本申请实施例还提供了一种图像遮蔽系统,包括:采集设备及处理设备;其中,
所述采集设备,用于采集视频流,将所采集的视频流发送给所述处理设备;
所述处理设备,用于接收所述采集设备发送的视频流,针对所接收到的视频流中的每一帧图像,检测所述图像中待遮蔽的目标区域;将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
可选的,所述系统中还包括:播放设备,
所述处理设备,还用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至所述播放设备;
所述播放设备,用于接收所述处理设备发送的所述封装后的图像;对所述封装后的图像进行解封装,得到遮蔽后的图像;播放所述遮蔽后的图像。
应用本申请所示实施例,检测每一帧图像中待遮蔽的目标区域,将该目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像,可见,本方案中,并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域(目标区域)在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果。
当然,实施本申请的任一产品或方法并不一定需要同时达到以上所述的所有优点。
附图说明
为了更清楚地说明本申请实施例和现有技术的技术方案,下面对实施例和现有技术中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的一种图像遮蔽方法的流程示意图;
图2为本申请实施例提供的图像坐标系示意图;
图3为本申请实施例提供的一种应用场景示意图;
图4为本申请实施例提供的一种图像遮蔽装置的结构示意图;
图5为本申请实施例提供的一种电子设备的结构示意图;
图6为本申请实施例提供的图像遮蔽系统的第一种结构示意图;
图7为本申请实施例提供的图像遮蔽系统的第二种结构示意图;
图8为本申请实施例提供的图像遮蔽系统的第三种结构示意图。
具体实施方式
为使本申请的目的、技术方案、及优点更加清楚明白,以下参照附图并举实施例,对本申请进一步详细说明。显然,所描述的实施例仅仅是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
为了解决上述技术问题,本申请实施例提供了一种图像遮蔽方法、装置、设备及系统。该方法可以应用于具有图像处理功能的设备,具体不做限定。
下面首先对本申请实施例提供的一种图像遮蔽方法进行详细说明。
图1为本申请实施例提供的一种图像遮蔽方法的流程示意图,包括:
S101:针对每一帧图像,检测该图像中待遮蔽的目标区域。
S102:将该目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
应用本申请图1所示实施例,检测每一帧图像中待遮蔽的目标区域,将该目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像,可见,本方案中,并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域(目标区域)在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果。
下面针对图1所示实施例进行详细说明:
S101:针对每一帧图像,检测该图像中待遮蔽的目标区域。
执行本方案的设备(或者说执行主体,以下简称电子设备)可以为图像采集设备,也可以为与图像采集设备通信连接的其他电子设备。如果电子设备为图像采集设备,则图像采集设备可以针对采集到的视频流中的每一帧图像,应用本方案进行图像遮蔽。如果电子设备为与图像采集设备通信连接的其他电子设备,则该电子设备可以从图像采集设备中获取其采集到的视频流,并针对该视频流中的每一帧图像,应用本方案进行图像遮蔽。
或者,应用本方案,也可以对单张图像进行遮蔽处理,具体不做限定。
作为一种实施方式,S101可以包括:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;根据所述输出结果,确定所述图像中待遮蔽的目标区域。
需要说明的是,上述“待遮蔽的图像内容”与S101中待遮蔽的目标区域中的内容一致。举例来说,如果上述“待遮蔽的图像内容”为人脸,则S101中待遮蔽的目标区域中的内容也为人脸;如果上述“待遮蔽的图像内容”为车牌,则S101中待遮蔽的目标区域中的内容也为车牌。
如果需要对图像中的人脸区域进行遮蔽,则预先利用深度学习算法,对人脸图像样本进行学习,得到人脸检测模型;S101中将图像输入至该人脸检测模型,根据输出结果可以确定图像中的人脸区域。
再举一例,如果需要对图像中的某个特定物体进行遮蔽,假设该物体为车牌,则预先利用深度学习算法,对车牌图像样本进行学习,得到车牌检测模型;S101中将图像输入至该车牌检测模型,根据输出结果可以确定图像中的车牌区域。
或者,也可以通过其他方式检测图像中待遮蔽的目标区域:
如果待遮蔽的目标区域为人脸,可以预先建立人脸特征点模型。具体的,可以提取鼻子、眼睛、嘴巴等部位的特征点,并根据面部比例,构建人脸特征点模型。将图像与该人脸特征点模型进行匹配,匹配成功的区域即为人脸区域,也就是待遮蔽的目标区域。
如果待遮蔽的目标区域为车牌,可以利用车牌识别技术识别图像中的车牌区域,等等,具体检测方式不做限定。
在一种实施方式中,执行S101后可以得到一个目标矩形框,该目标矩形框中的图像区域即为目标区域。
可以以图像的某个角点为原点构建坐标系,举例来说,如图2所示,以图 像左上角的角点作为原点构建坐标系,确定目标矩形框的四个角点的坐标分别为:A(W/3,H/3)、B(2W/3,H/3)、C(W/3,2H/3)、D(2W/3,2H/3),将坐标位于W/3-2W/3,H/3-2H/3区间的像素点作为目标区域中的像素点进行后续处理。
S102:将该目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
可以理解,图像有多种格式,比如RGB(红绿蓝)格式、或者YUV(亮度和色差)格式等。如果本实施例的图像数据为RGB格式,则可以将目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值。
可以理解,将目标区域中的每个像素点的RGB值都调整为相同的值,则该目标区域为一片色彩相同的区域,这样,目标区域的原画面内容被遮蔽,得到了遮蔽后的图像。
需要说明的是,这里所说的“将目标区域中的每个像素点的RGB值都调整为相同的值”并不是说一个像素点的R值、G值、B值相同,而是说,每个像素点的R值均相同,G值均相同,B值均相同。举例来说,将目标区域中的每个像素点的R值都调整为100,G值都调整为120,B值都调整为130,则目标区域的原画面内容被遮蔽。
或者,也可以将目标区域中的每个像素点的RGB值全部置0,或者全部置为255,或者全部置为某一数值,这样,一个像素点的R值、G值、B值相同,每个像素点的R值均相同,G值均相同,B值均相同,也可以遮蔽原画面内容。
也就是说,该第一组预设值中包含预设R值、预设G值、预设B值,这三个数值可以相同或不同,这三个数值可以为0-255中任意值;将目标区域中的每个像素点的R值都调整为第一组预设值中的预设R值,将目标区域中的每个像素点的G值都调整为第一组预设值中的预设G值,将目标区域中的每个像素点的B值都调整为第一组预设值中的预设B值,便可以达到遮蔽图像内容的效果。
或者,也可以设定一组数值区间,该数值区间中包含预设R值区间、预设G值区间、预设B值区间;将目标区域中的每个像素点的R值都调整至该预设R值区间,将目标区域中的每个像素点的G值都调整至该预设G值区间,将目标区域中的每个像素点的B值都调整至该预设B值区间。
可以理解,将目标区域中的每个像素点的RGB值都调整至同一组数值区间,也就是将每个像素点的RGB值都调整为近似的值,这样该目标区域为一片色彩相近的区域,目标区域的原画面内容被遮蔽,得到了遮蔽后的图像。
如果本实施例的图像数据为YUV格式,则可以将目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
可以根据经验值预先设定该第二组预设值,执行S102时,直接读取该第二组预设值,得到该预设Y值、预设U值、预设V值,将目标区域中的每个像素点的Y值置为该预设Y值,将目标区域中的每个像素点的U值置为该预设U值,将目标区域中的每个像素点的V值置为该预设V值。
或者,也可以采用其他方式设定该第二组预设值,比如,可以对图像的YUV值进行反复调整,当调整至遮蔽效果较佳时,记录该次调整的YUV值,作为该第二组预设值。
或者,也可以采用如下步骤确定所述第二组预设值:
确定一组RGB预设值;根据RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
确定一组RGB预设值,可以随机确定,比如,确定R值=100,G值=120,B值=130,则利用RGB格式数据与YUV格式数据的转换关系,将RGB100、120、130转换为YUV格式下的值,也就是该第二组预设值。
具体转换方式有多种,比如,查表法、整型算法等等,具体不做限定。下面介绍一种转换算式:
R=Y+1.4075*(V-128)
G=Y–0.3455*(U–128)–0.7169*(V–128)
B=Y+1.779*(U–128)
将R值=100,G值=120,B值=130代入上述算式,可以计算得到YUV的值,将得到的该YUV的值作为第二组预设值。
作为一种实施方式,S102中的预设遮蔽数据也可以为马赛克数据,也就是将目标区域中的图像数据置换为马赛克数据。
可以对视频流中的每一帧图像都应用上述方案:自动检测目标区域,自动修改目标区域的图像数据,这样,在处理一段视频的过程中,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果。
应用本申请图1所示实施例,第一方面,本方案中,并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域(目标区域)在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果;第二方面,如果只是在图像展示过程中设置遮蔽图片,传输过程中的图像仍为未遮蔽的图像,仍有可能泄露图像中的隐私内容,安全性较低,而本方案对图像数据本身进行修改,传输过程中的图像为遮蔽后的图像,提高了数据的安全性。
作为一种实施方式,在S102之后,可以对该遮蔽后的图像进行展示,或者,也可以利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
在本实施方式中,电子设备自身可以对遮蔽后的图像进行展示,也可以将遮蔽后的图像编码封装后发送给播放器进行展示。
举个例子,假设获取的图像格式为YUV格式,则将目标区域中每个像素点的YUV值修改为第二组预设值后,可以将修改后的YUV格式的数据编码成H.264格式,再将编码后的数据封装成PS(Program Stream)数据,将该 PS数据发送至播放器进行播放。
或者,作为另一种实施方式,电子设备自身先对遮蔽后的图像进行展示,判断所展示图像的遮蔽效果是否合格;如果是,利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
举例来说,电子设备可以将S102中得到的遮蔽后的图像展示给用户,如果用户确定所展示图像的遮蔽效果合格,则向电子设备发送第一指令,该第一指令可以为确认指令、播放指令等。电子设备如果接收到该第一指令,则利用预设编码封装格式,将S102中得到的遮蔽后的图像进行编码封装,得到封装后的图像;将封装后的图像发送至播放器进行播放。
在此基础上,如果用户确定所展示图像的遮蔽效果不合格,还可以向电子设备发送第二指令,该第二指令可以为停止指令等。电子设备接收到该第二指令后,可以停止对视频流中的其他图像进行遮蔽处理。
再举个例子,电子设备将S102中得到的遮蔽后的图像进行展示,电子设备可以对所展示的图像作进一步的判断,判断所展示图像中是否存在需要遮蔽的敏感区域,如果不存在,则表示所展示图像的遮蔽效果合格。这种情况下,电子设备利用预设编码封装格式,将S102中得到的遮蔽后的图像进行编码封装,得到封装后的图像;将封装后的图像发送至播放器进行播放。
进一步判断所展示图像中是否存在需要遮蔽的敏感区域,可以利用上述检测模型进行检测,或者,如果该敏感区域为人脸,可以利用人脸特征点模型进行匹配,如果在遮蔽后的图像中检测出了人脸区域,则判定遮蔽效果不合格。或者,如果该敏感区域为车牌,可以利用车牌识别技术判断图像中是否存在车牌号,如果在遮蔽后的图像中检测出了车牌区域,则判定遮蔽效果不合格。
下面结合图3,介绍一个具体的实施例:
1、采集设备实时采集视频流;采集设备通过交换设备将该视频流发送给处理设备;或者,采集设备也可以直接将该视频流发送给处理设备。
2、处理设备接收该视频流,对该视频流进行解码得到YUV数据。
3、处理设备针对YUV数据中的每一帧图像,将该图像输入至预先学习得到的检测模型中,输出一个目标矩形框,目标矩形框中的图像区域即为待遮蔽的目标区域。
4、处理设备将目标区域中每个像素点的YUV值置为第二组预设值。
预先确定该第二组预设值的过程可以包括:确定一组RGB预设值;利用RGB格式数据与YUV格式数据的转换关系,反解像素点的YUV值,求解得到的值即为第二组预设值。
5、处理设备将经过步骤4处理的图像作为遮蔽后的图像进行展示,如果所展示图像的遮蔽效果合格,处理设备利用预设编码封装格式,对该遮蔽后的图像进行编码封装,得到封装后的图像。
6、处理设备将该封装后的图像发送至播放器进行播放。该播放器可以为处理设备中内置的播放器,也可以为其他播放设备,具体不做限定。
由此可见,应用上述方案,第一方面,并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域(目标区域)在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果;第二方面,如果只是在图像展示过程中设置遮蔽图片,传输过程中的图像仍为未遮蔽的图像,仍有可能泄露图像中的隐私内容,安全性较低,而本方案对图像数据本身进行修改,传输过程中的图像为遮蔽后的图像,提高了数据的安全性;第三方面,由处理设备对图像进行遮蔽处理,相比于采集设备处理图像,降低了采集设备的成本。
与上述方法实施例相对应,本申请实施例还提供一种图像遮蔽装置。
图4为本申请实施例提供的一种图像遮蔽装置的结构示意图,包括:
检测模块401,用于针对每一帧图像,检测所述图像中待遮蔽的目标区域;
修改模块402,用于将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
作为一种实施方式,检测模块401,具体可以用于:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;根据所述输出结果,确定所述图像中待遮蔽的目标区域。
作为一种实施方式,修改模块402,具体可以用于:
在所述图像数据为RGB格式的情况下,将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
在所述图像数据为YUV格式的情况下,将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
作为一种实施方式,所述装置还可以包括:
确定模块(图中未示出),用于确定一组RGB预设值;利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
作为一种实施方式,所述装置还可以包括:展示模块(图中未示出),或者还包括:封装模块和发送模块(图中未示出),其中,
展示模块,用于对所述遮蔽后的图像进行展示;
封装模块,用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;
发送模块,用于将所述封装后的图像发送至播放器进行播放。
作为一种实施方式,所述装置还可以包括:展示模块、判断模块、封装模块和发送模块(图中未示出),其中,
展示模块,用于对所述遮蔽后的图像进行展示;
判断模块,用于判断所展示图像的遮蔽效果是否合格;如果是,触发封装模块;
封装模块,用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;
发送模块,用于将所述封装后的图像发送至播放器进行播放。
应用本申请图4所示实施例,第一方面,并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域(目标区域)在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果;第二方面,如果只是在图像展示过程中设置遮蔽图片,传输过程中的图像仍为未遮蔽的图像,仍有可能泄露图像中的隐私内容,安全性较低,而本方案对图像数据本身进行修改,传输过程中的图像为遮蔽后的图像,提高了数据的安全性;第三方面,由处理设备对图像进行遮蔽处理,相比于采集设备处理图像,降低了采集设备的成本。
本申请实施例还提供了一种电子设备,图5为本申请实施例提供的一种电子设备的结构示意图,包括:处理器501和存储器502,其中,存储器502用于存储可执行程序代码;处理器501通过读取存储器502中存储的可执行程序代码来运行与可执行程序代码对应的程序,以用于执行以下步骤:
针对每一帧图像,检测所述图像中待遮蔽的目标区域;
将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
作为一种实施方式,处理器501还可以用于执行以下步骤:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;
根据所述输出结果,确定所述图像中待遮蔽的目标区域。
作为一种实施方式,处理器501还可以用于执行以下步骤:
在所述图像数据为RGB格式的情况下,将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含 预设R值、预设G值、预设B值;
在所述图像数据为YUV格式的情况下,将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
作为一种实施方式,处理器501还可以用于执行以下步骤:
确定一组RGB预设值;
利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
作为一种实施方式,处理器501还可以用于执行以下步骤:
在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,对所述遮蔽后的图像进行展示;
或者,
利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
作为一种实施方式,处理器501还可以用于执行以下步骤:
在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,对所述遮蔽后的图像进行展示;
判断所展示图像的遮蔽效果是否合格;
如果是,利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
上述电子设备提到的存储器可以包括随机存取存储器(Random Access Memory,RAM),也可以包括非易失性存储器(Non-Volatile Memory,NVM),例如至少一个磁盘存储器。可选的,存储器还可以是至少一个位于远离前述处理器的存储装置。
上述的处理器可以是通用处理器,包括中央处理器(Central Processing Unit,CPU)、网络处理器(Network Processor,NP)等;还可以是数字信号处理器(Digital Signal Processing,DSP)、专用集成电路(Application Specific Integrated Circuit,ASIC)、现场可编程门阵列(Field-Programmable Gate Array,FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。
本申请实施例还提供一种计算机可读存储介质,所述计算机可读存储介质内存储有计算机程序,所述计算机程序被处理器执行时实现如下步骤:
针对每一帧图像,检测所述图像中待遮蔽的目标区域;
将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
作为一种实施方式,所述计算机程序被处理器执行时还可以实现如下步骤:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;
根据所述输出结果,确定所述图像中待遮蔽的目标区域。
作为一种实施方式,所述计算机程序被处理器执行时还可以实现如下步骤:
在所述图像数据为RGB格式的情况下,将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
在所述图像数据为YUV格式的情况下,将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
作为一种实施方式,所述计算机程序被处理器执行时还可以实现如下步骤:
确定一组RGB预设值;
利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
作为一种实施方式,所述计算机程序被处理器执行时还可以实现如下步骤:
在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,对所述遮蔽后的图像进行展示;
或者,
利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
作为一种实施方式,所述计算机程序被处理器执行时还可以实现如下步骤:
在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,对所述遮蔽后的图像进行展示;
判断所展示图像的遮蔽效果是否合格;
如果是,利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
本申请实施例还提供一种可执行程序代码,所述可执行程序代码用于被运行以执行上述任一种图像遮蔽方法。
本申请实施例还提供一种图像遮蔽系统,如图6所示,包括:采集设备及处理设备;其中,
采集设备,用于采集视频流,将所采集的视频流发送给所述处理设备;
处理设备,用于接收所述采集设备发送的视频流,针对所接收到的视频流中的每一帧图像,检测所述图像中待遮蔽的目标区域;将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
作为一种实施方式,该系统还可以如图7所示,还包括播放设备,在本实施方式中,处理设备,还可以用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至所述播放设备;
播放设备,用于接收所述处理设备发送的所述封装后的图像;对所述封装后的图像进行解封装,得到遮蔽后的图像;播放所述遮蔽后的图像。
或者,作为另一种实施方式,如图8所示,处理设备中可以内置播放器,处理设备可以利用预设编码封装格式,对该遮蔽后的图像进行编码封装,得到封装后的图像;将该封装后的图像发送至该内置的播放器进行播放;该播放器对所述封装后的图像进行解封装,得到遮蔽后的图像;播放所述遮蔽后的图像。
作为一种实施方式,该处理设备还可以用于:
将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;根据所述输出结果,确定所述图像中待遮蔽的目标区域。
作为一种实施方式,该处理设备还可以用于:
在所述图像数据为RGB格式的情况下,将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
在所述图像数据为YUV格式的情况下,将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
作为一种实施方式,该处理设备还可以用于:
确定一组RGB预设值;
利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
作为一种实施方式,该处理设备还可以用于:
在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,对所述遮蔽后的图像进行展示;
或者,
利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
作为一种实施方式,该处理设备还可以用于:
在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,对所述遮蔽后的图像进行展示;
判断所展示图像的遮蔽效果是否合格;
如果是,利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
应用本申请所示实施例,第一方面,并不是基于固定位置的遮蔽图片遮蔽图像中的敏感区域,而是先检测出敏感区域(目标区域)在图像中的位置,再对该位置处的图像数据进行修改,即使敏感区域发生了移动,仍然能够检测出敏感区域,并对其进行遮蔽,提高了遮蔽效果;第二方面,如果只是在图像展示过程中设置遮蔽图片,传输过程中的图像仍为未遮蔽的图像,仍有可能泄露图像中的隐私内容,安全性较低,而本方案对图像数据本身进行修改,传输过程中的图像为遮蔽后的图像,提高了数据的安全性;第三方面,由处理设备对图像进行遮蔽处理,相比于采集设备处理图像,降低了采集设备的成本。
需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列 出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
本说明书中的各个实施例均采用相关的方式描述,各个实施例之间相同相似的部分互相参见即可,每个实施例重点说明的都是与其他实施例的不同之处。尤其,对于图4所示的图像遮蔽装置实施例、图5所示的电子设备实施例、图6-8所示的系统实施例、上述计算机可读存储介质实施例、以及上述可执行程序代码实施例而言,由于其基本相似于图1-3所示的图像遮蔽方法实施例,所以描述的比较简单,相关之处参见图1-3所示的图像遮蔽方法实施例的部分说明即可。
以上所述仅为本申请的较佳实施例而已,并不用以限制本申请,凡在本申请的精神和原则之内,所做的任何修改、等同替换、改进等,均应包含在本申请保护的范围之内。

Claims (17)

  1. 一种图像遮蔽方法,其特征在于,包括:
    针对每一帧图像,检测所述图像中待遮蔽的目标区域;
    将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
  2. 根据权利要求1所述的方法,其特征在于,所述检测所述图像中待遮蔽的目标区域,包括:
    将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;
    根据所述输出结果,确定所述图像中待遮蔽的目标区域。
  3. 根据权利要求1所述的方法,其特征在于,在所述图像数据为RGB格式的情况下,所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像,包括:
    将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
    在所述图像数据为YUV格式的情况下,所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像,包括:
    将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
  4. 根据权利要求3所述的方法,其特征在于,采用如下步骤确定所述第二组预设值:
    确定一组RGB预设值;
    利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
  5. 根据权利要求1所述的方法,其特征在于,在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,还包括:
    对所述遮蔽后的图像进行展示;
    或者,
    利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
  6. 根据权利要求1所述的方法,其特征在于,在所述将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像之后,还包括:
    对所述遮蔽后的图像进行展示;
    判断所展示图像的遮蔽效果是否合格;
    如果是,利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至播放器进行播放。
  7. 一种图像遮蔽装置,其特征在于,包括:
    检测模块,用于针对每一帧图像,检测所述图像中待遮蔽的目标区域;
    修改模块,用于将所述目标区域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
  8. 根据权利要求7所述的装置,其特征在于,所述检测模块,具体用于:
    将所述图像输入至预先学习得到的检测模型中,得到输出结果;其中,所述检测模型为:利用深度学习算法,对预设图像样本进行学习得到的,所述预设图像样本中包含待遮蔽的图像内容;根据所述输出结果,确定所述图像中待遮蔽的目标区域。
  9. 根据权利要求7所述的装置,其特征在于,所述修改模块,具体用于:
    在所述图像数据为RGB格式的情况下,将所述目标区域中的每个像素点的RGB值置为第一组预设值,得到遮蔽后的图像,所述第一组预设值中包含预设R值、预设G值、预设B值;
    在所述图像数据为YUV格式的情况下,将所述目标区域中的每个像素点的YUV值置为第二组预设值,得到遮蔽后的图像,所述第二组预设值包含预设Y值、预设U值、预设V值。
  10. 根据权利要求9所述的装置,其特征在于,所述装置还包括:
    确定模块,用于确定一组RGB预设值;利用RGB格式数据与YUV格式数据的转换关系,将所述一组RGB预设值转换为第二组预设值。
  11. 根据权利要求7所述的装置,其特征在于,所述装置还包括:展示模块,或者还包括:封装模块和发送模块,其中,
    展示模块,用于对所述遮蔽后的图像进行展示;
    封装模块,用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;
    发送模块,用于将所述封装后的图像发送至播放器进行播放。
  12. 根据权利要求7所述的装置,其特征在于,所述装置还包括:
    展示模块,用于对所述遮蔽后的图像进行展示;
    判断模块,用于判断所展示图像的遮蔽效果是否合格;如果是,触发封装模块;
    封装模块,用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;
    发送模块,用于将所述封装后的图像发送至播放器进行播放。
  13. 一种电子设备,其特征在于,包括处理器和存储器,其中,存储器用于存储可执行程序代码,处理器通过读取存储器中存储的可执行程序代码来运行与可执行程序代码对应的程序,以用于执行权利要求1-6任一项图像遮蔽方法。
  14. 一种图像遮蔽系统,其特征在于,包括:采集设备及处理设备;其中,
    所述采集设备,用于采集视频流,将所采集的视频流发送给所述处理设备;
    所述处理设备,用于接收所述采集设备发送的视频流,针对所接收到的视频流中的每一帧图像,检测所述图像中待遮蔽的目标区域;将所述目标区 域中的图像数据修改为预设遮蔽数据,得到遮蔽后的图像。
  15. 根据权利要求14所述的系统,其特征在于,所述系统中还包括:播放设备,
    所述处理设备,还用于利用预设编码封装格式,对所述遮蔽后的图像进行编码封装,得到封装后的图像;将所述封装后的图像发送至所述播放设备;
    所述播放设备,用于接收所述处理设备发送的所述封装后的图像;对所述封装后的图像进行解封装,得到遮蔽后的图像;播放所述遮蔽后的图像。
  16. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质内存储有计算机程序,所述计算机程序被处理器执行时实现权利要求1-6任一项图像遮蔽方法。
  17. 一种可执行程序代码,其特征在于,所述可执行程序代码用于被运行以执行权利要求1-6任一项图像遮蔽方法。
PCT/CN2018/102467 2017-08-29 2018-08-27 一种图像遮蔽方法、装置、设备及系统 WO2019042243A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710754705.6 2017-08-29
CN201710754705.6A CN109427082A (zh) 2017-08-29 2017-08-29 一种图像遮蔽方法、装置、设备及系统

Publications (1)

Publication Number Publication Date
WO2019042243A1 true WO2019042243A1 (zh) 2019-03-07

Family

ID=65503412

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/102467 WO2019042243A1 (zh) 2017-08-29 2018-08-27 一种图像遮蔽方法、装置、设备及系统

Country Status (2)

Country Link
CN (1) CN109427082A (zh)
WO (1) WO2019042243A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110473149B (zh) * 2019-03-13 2020-06-09 北方工业大学 关键大数据模糊化处理系统
CN110956578B (zh) * 2019-03-13 2020-09-01 深圳市中壬银兴信息技术有限公司 关键大数据模糊化处理方法
CN109977937B (zh) * 2019-03-26 2020-11-03 北京字节跳动网络技术有限公司 图像处理方法、装置及设备
CN110290288B (zh) * 2019-06-03 2022-01-04 Oppo广东移动通信有限公司 图像处理器、图像处理方法、拍摄装置和电子设备
TWI770410B (zh) * 2019-08-12 2022-07-11 華南商業銀行股份有限公司 影像辨識系統

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101742228A (zh) * 2008-11-19 2010-06-16 新奥特硅谷视频技术有限责任公司 一种应用于数字法庭中的预处理方法和系统
CN102306290A (zh) * 2011-10-14 2012-01-04 刘伟华 一种基于视频的人脸跟踪识别技术
CN104270580A (zh) * 2014-09-28 2015-01-07 济南大学 一种视频特效快速实现方法
US20170147892A1 (en) * 2015-11-20 2017-05-25 Panasonic Intellectual Property Corporation Of America Method for processing image and computer-readable non-transitory recording medium storing program
CN106851332A (zh) * 2017-01-04 2017-06-13 北京百度网讯科技有限公司 视频流处理方法、装置和系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103049755B (zh) * 2012-12-28 2016-08-10 合一网络技术(北京)有限公司 一种实现动态视频马赛克的方法及装置
CN105357466A (zh) * 2015-11-20 2016-02-24 小米科技有限责任公司 视频通信方法及装置
CN105957001A (zh) * 2016-04-18 2016-09-21 深圳感官密码科技有限公司 一种隐私保护方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101742228A (zh) * 2008-11-19 2010-06-16 新奥特硅谷视频技术有限责任公司 一种应用于数字法庭中的预处理方法和系统
CN102306290A (zh) * 2011-10-14 2012-01-04 刘伟华 一种基于视频的人脸跟踪识别技术
CN104270580A (zh) * 2014-09-28 2015-01-07 济南大学 一种视频特效快速实现方法
US20170147892A1 (en) * 2015-11-20 2017-05-25 Panasonic Intellectual Property Corporation Of America Method for processing image and computer-readable non-transitory recording medium storing program
CN106851332A (zh) * 2017-01-04 2017-06-13 北京百度网讯科技有限公司 视频流处理方法、装置和系统

Also Published As

Publication number Publication date
CN109427082A (zh) 2019-03-05

Similar Documents

Publication Publication Date Title
WO2019042243A1 (zh) 一种图像遮蔽方法、装置、设备及系统
WO2022179108A1 (zh) 投影校正方法、装置、存储介质和电子设备
US10163027B2 (en) Apparatus for and method of processing image based on object region
CN109166156B (zh) 一种摄像头标定图像的生成方法、移动终端及存储介质
US11004214B2 (en) Image processing apparatus, image processing method, and storage medium
CN110263621B (zh) 图像识别方法、装置及可读存储介质
TWI419082B (zh) 偵測視訊序列中的移動物體影像的方法以及影像處理系統
CN110796600B (zh) 一种图像超分重建方法、图像超分重建装置及电子设备
US20200126193A1 (en) Method and device for video processing, electronic device, and storage medium
US20180234665A1 (en) On screen display (osd) information generation camera, osd information synthesis terminal, and osd information sharing system including the same
US11955032B2 (en) Sensor device and encryption method
US10863113B2 (en) Image processing apparatus, image processing method, and storage medium
CN113709519B (zh) 一种用于确定直播遮挡区域的方法与设备
CN110674729A (zh) 基于热能估算的人数识别方法、计算机装置及计算机可读存储介质
CN110944227A (zh) 目标隐私保护方法、装置、硬盘录像机及存储介质
CN108712400A (zh) 数据传输方法、装置、计算机可读存储介质和电子设备
TWI502988B (zh) 電子裝置及增強影像解析度之方法及其電腦可讀取媒體
WO2019109722A1 (zh) 隐私遮蔽处理方法、装置、电子设备及存储介质
CN109726613B (zh) 一种用于检测的方法和装置
EP3920461A1 (en) Sensor device and encryption method
US20140307116A1 (en) Method and system for managing video recording and/or picture taking in a restricted environment
CN104851114B (zh) 一种实现图像局部变色的方法及终端
CN112241735A (zh) 一种图像处理方法、装置及系统
JP2016046701A (ja) 画像処理装置、画像処理方法、及び、プログラム
KR101080066B1 (ko) 차량 번호판 식별 방법 및 차량 번호판 식별 장치

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: 18849598

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: 18849598

Country of ref document: EP

Kind code of ref document: A1