WO2021004261A1 - 深度数据的滤波方法、装置、电子设备和可读存储介质 - Google Patents

深度数据的滤波方法、装置、电子设备和可读存储介质 Download PDF

Info

Publication number
WO2021004261A1
WO2021004261A1 PCT/CN2020/097464 CN2020097464W WO2021004261A1 WO 2021004261 A1 WO2021004261 A1 WO 2021004261A1 CN 2020097464 W CN2020097464 W CN 2020097464W WO 2021004261 A1 WO2021004261 A1 WO 2021004261A1
Authority
WO
WIPO (PCT)
Prior art keywords
depth
pixel
preset
change area
environment change
Prior art date
Application number
PCT/CN2020/097464
Other languages
English (en)
French (fr)
Inventor
康健
Original Assignee
Oppo广东移动通信有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Oppo广东移动通信有限公司 filed Critical Oppo广东移动通信有限公司
Priority to EP20837151.8A priority Critical patent/EP3975107A4/en
Publication of WO2021004261A1 publication Critical patent/WO2021004261A1/zh
Priority to US17/559,103 priority patent/US20220114744A1/en

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/22Matching criteria, e.g. proximity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00Image enhancement or restoration
    • G06T5/70Denoising; Smoothing
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/50Depth or shape recovery
    • G06T7/55Depth or shape recovery from multiple images
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical techniques
    • G01B11/22Measuring arrangements characterised by the use of optical techniques for measuring depth
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/88Lidar systems specially adapted for specific applications
    • G01S17/89Lidar systems specially adapted for specific applications for mapping or imaging
    • G01S17/8943D imaging with simultaneous measurement of time-of-flight at a 2D array of receiver pixels, e.g. time-of-flight cameras or flash lidar
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S7/00Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/48Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
    • G01S7/483Details of pulse systems
    • G01S7/486Receivers
    • G01S7/4865Time delay measurement, e.g. time-of-flight measurement, time of arrival measurement or determining the exact position of a peak
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/22Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • G06V10/225Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition based on a marking or identifier characterising the area
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/30Noise filtering
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/34Smoothing or thinning of the pattern; Morphological operations; Skeletonisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/70Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/72Data preparation, e.g. statistical preprocessing of image or video features
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/70Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/761Proximity, similarity or dissimilarity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10016Video; Image sequence
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10028Range image; Depth image; 3D point clouds
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20172Image enhancement details
    • G06T2207/20182Noise reduction or smoothing in the temporal domain; Spatio-temporal filtering

Definitions

  • ToF Time of Flight
  • sensors determine the distance between the sensor and the object by calculating the flight time of the pulse signal. Due to various uncertainties in the measurement process, various errors are brought, and these errors have a lot of The large randomness causes the ToF depth measurement error within the measurement range to be about 1%.
  • the above measurement error can be accepted, but it is hoped that the sensor can achieve time consistency within a limited time.
  • the embodiments of the present application provide a filtering method, device, electronic device, and readable storage medium for depth data.
  • An embodiment of the first aspect of the present application proposes a method for filtering depth data, including:
  • the area formed by each pixel with the depth difference less than the preset absolute depth error is marked as the first environment change area; wherein, the depth value of each pixel point in the previous frame depth map and the preset error percentage are determined The preset absolute depth error;
  • An embodiment of the second aspect of the present application proposes a depth data filtering device, including:
  • the acquisition module is used to acquire the depth difference of each pixel point between two consecutive depth maps
  • the first marking module is used to mark the area formed by the pixels whose depth difference is less than the preset absolute depth error as the first environment change area; wherein, according to the depth value and preset value of each pixel in the previous frame depth map Set the error percentage to determine the preset absolute depth error;
  • the second generating module is configured to generate a second pixel corresponding to each pixel in the second environment change area according to the depth difference, the preset original smoothing coefficient after the reduction process, and the pixel depth error value of the current frame. Similarity weight;
  • the processing module is configured to perform filtering processing on the first environmental change area according to the first similarity weight, and perform filtering processing on the second environmental change area with the second similarity weight.
  • the embodiment of the third aspect of the present application proposes an electronic device, an image sensor, a memory, a processor, and a computer program stored on the memory and running on the processor, the image sensor is electrically connected to the processor, and When the processor executes the program, it implements the depth data filtering method described in the foregoing method embodiment.
  • the embodiment of the fourth aspect of the present application proposes a computer-readable storage medium on which a computer program is stored.
  • the program is executed by a processor, the method for filtering depth data as described in the foregoing method embodiment is implemented.
  • FIG. 1 is a schematic flowchart of a depth acquisition method provided by an embodiment of this application.
  • FIG. 2 is a schematic flowchart of a method for filtering depth data according to an embodiment of the application
  • FIG. 3 is a schematic diagram of obtaining an original depth value according to an embodiment of the application.
  • FIG. 4 is a schematic flowchart of another depth data filtering method provided by an embodiment of the application.
  • Fig. 5 is a schematic structural diagram of a depth data filtering device according to an embodiment of the present application.
  • FIG. 2 is a schematic flowchart of a method for filtering depth data according to an embodiment of the application. As shown in Figure 2, the method includes the following steps:
  • Step 102 Mark an area formed by each pixel with a depth difference less than a preset absolute depth error as a first environment change area, where it is determined according to the depth value of each pixel in the previous frame depth map and the preset error percentage The absolute depth error is preset.
  • Step 103 Mark an area formed by pixels with a depth difference greater than or equal to a preset absolute depth error as a second environment change area.
  • Step 104 Generate a first similarity weight corresponding to each pixel in the first environment change area according to the depth difference, the preset original smoothing coefficient after the enlargement process, and the pixel depth error value of the current frame.
  • Step 105 Generate a second similarity weight corresponding to each pixel in the second environment change area according to the depth difference, the preset original smoothing coefficient after the reduction process, and the depth error value of the pixel point of the current frame.
  • Step 106 Perform filtering processing on the first environmental change area according to the first similarity weight, and perform filtering processing on the second environmental change area according to the second similarity weight.
  • FIG. 4 is a schematic flowchart of another depth data filtering method provided by an embodiment of the application. As shown in Figure 4, the method includes the following steps:
  • Step 201 Obtain the depth difference of each pixel point between two consecutive depth maps.
  • Step 202 Mark an area formed by pixels with a depth difference smaller than a preset absolute depth error as a first environment change area, and mark the first area mask corresponding to the first environment change area.
  • Step 203 Mark an area formed by pixels whose depth difference is greater than or equal to a preset absolute depth error as a second environment change area, and mark the second area mask corresponding to the second environment change area.
  • Step 205 Obtain the first original depth value of the previous frame and the first original depth value of the current frame corresponding to each pixel in the first environment change area in the preset coordinate system, and compare the first similarity weight with the first original depth value of the previous frame. The product of an original depth value and the product of the third similarity weight and the first original depth value of the current frame are added together to obtain the first current frame depth value corresponding to each pixel in the first environment change area.
  • Step 206 Generate a second similarity weight corresponding to each pixel in the second environment change area according to the depth difference, the preset original smoothing coefficient after the reduction process, and the depth error value of the pixel point of the current frame.
  • w1 is s is the preset original smoothing coefficient
  • diff is the depth difference value, which represents the reflectivity difference between the previous and next frames
  • is the depth error value of the pixel point of the current frame.
  • the depth data filtering device includes: an acquisition module 501, a first marking module 502, a second marking module 503, and a first generation Module 504, second generation module 505 and processing module 506.
  • the acquiring module 501 is used to acquire the depth difference of each pixel point between two consecutive depth maps.
  • the first marking module 502 is used to mark the area formed by each pixel with a depth difference less than the preset absolute depth error as the first environment change area; wherein, according to the depth value of each pixel in the previous frame depth map and The preset error percentage determines the preset absolute depth error.
  • the first generating module 504 is configured to generate a first similarity weight corresponding to each pixel in the first environment change area according to the depth difference, the preset original smoothing coefficient after the amplification process, and the current frame pixel depth error value.
  • the processing module 506 is configured to perform filtering processing on the first environmental change area according to the first similarity weight, and perform filtering processing on the second environmental change area according to the second similarity weight.
  • a preset formula is applied to generate a similarity weight based on the depth difference of each pixel, the preset original smoothing coefficient, and the current frame pixel depth error value.
  • the preset formula is: Wherein, s is a preset original smoothing coefficient, diff is the depth difference value, and ⁇ is the pixel depth error value of the current frame.
  • this application also proposes an electronic device, including a memory, a processor, and a computer program stored on the memory and capable of running on the processor.
  • the processor executes the computer program, it realizes the Filtering method of depth data.
  • the embodiment of the present application also proposes a computer-readable storage medium on which a computer program is stored.
  • the program is executed by a processor, the method for filtering depth data as described in the foregoing method embodiment is implemented.
  • the ToF sensor determines the distance between the sensor and the object by calculating the flight time of the pulse signal, such as Where d is the depth, c is the speed of light, and t is the flight time; divided by 2 is because the pulse signal flies twice between the sensor and the object. Based on the description of the above background technology, it can be understood that the time for the ToF depth data is consistent sexual filtering is very important.
  • the method of acquiring the depth of each frame of ToF is shown in Figure 1.
  • the ToF sensor emits a modulated pulse signal
  • the surface of the object to be measured receives the pulse signal and reflects the signal
  • the ToF sensor receives the reflected signal
  • Decode the multi-frequency phase map then perform error correction on the ToF data according to the calibration parameters, then de-alias the multi-frequency signal, and convert the depth value from the radial coordinate system to the Cartesian coordinate system, and finally the depth value is time consistent Sexual filtering, which outputs relatively smooth depth results in the time dimension.
  • the above method is to filter all pixels in the full frame, which will result in insufficient smoothness of time-consistent filtering, and cause the technical problem of large jitter of depth data in the time domain.
  • Step 101 Obtain the depth difference of each pixel point between two consecutive depth maps.
  • Step 102 Mark an area formed by each pixel with a depth difference less than a preset absolute depth error as a first environment change area, where it is determined according to the depth value of each pixel in the previous frame depth map and the preset error percentage The absolute depth error is preset.
  • Step 103 Mark an area formed by pixels with a depth difference greater than or equal to a preset absolute depth error as a second environment change area.
  • the area formed by the pixels whose depth difference is less than the preset absolute depth error is marked as the first environment change area, and the area formed by the pixels whose depth difference is greater than or equal to the preset absolute depth error The area is marked as the second environmental change area.
  • Step 105 Generate a second similarity weight corresponding to each pixel in the second environment change area according to the depth difference, the preset original smoothing coefficient after the reduction process, and the depth error value of the pixel point of the current frame.
  • Step 106 Perform filtering processing on the first environmental change area according to the first similarity weight, and perform filtering processing on the second environmental change area according to the second similarity weight.
  • the preset original smoothing coefficient after zooming and the pixel of the current frame The depth error value generates the first similarity weight corresponding to each pixel in the first environment change area and generates the first similarity weight according to the depth difference of each pixel, the preset original smoothing coefficient after reduction processing, and the depth error value of the current frame pixel.
  • the second similarity weight corresponding to each pixel in the second environment change area is necessary to perform smoothing processing in different areas.
  • the first environment change area which is the environment slowly changing area, has relatively high smoothness, that is, relatively high credibility, and the preset original smoothing coefficient needs to be enlarged to increase the first similarity weight.
  • the first similarity corresponding to each pixel in the first environmental change area can be generated according to the above formula according to the depth difference of each pixel, the preset original smoothing coefficient after the amplification process, and the depth error of the current frame pixel. Degree weight.
  • the second environmental change area is the rapid environmental change area, which has relatively low smoothness, that is, relatively low credibility, and the preset original smoothing coefficient needs to be reduced to weight the second similarity.
  • the preset original smoothing coefficient is an original empirical value set according to the temporal consistency filter.
  • the depth data filtering method of the embodiment of the present application obtains the depth difference between each pixel in two consecutive frames of depth maps; the depth difference is smaller than the preset absolute depth error.
  • the area is marked as the first environment change area; among them, the preset absolute depth error is determined according to the depth value of the depth map of each pixel in the previous frame and the preset error percentage; each pixel whose depth difference is greater than or equal to the preset absolute depth error.
  • the area formed by the dots is marked as the second environment change area; according to the depth difference, the preset original smoothing coefficient after the amplification process, and the pixel depth error value of the current frame, the first environment change area corresponding to each pixel in the first environment change area is generated.
  • Step 203 Mark an area formed by pixels whose depth difference is greater than or equal to a preset absolute depth error as a second environment change area, and mark the second area mask corresponding to the second environment change area.
  • the preset absolute depth error can be selected and set as needed.
  • the depth value of a pixel before and after the frame is [500, 502]
  • the preset absolute depth error is one hundredth, which is 5, and the pixel is currently
  • the depth difference between the depth map of the frame and the depth map of the previous frame is 2, which is less than 5, and is correspondingly marked as the first environment change area; for another example, the depth value of a pixel before and after the frame is [500, 520],
  • the preset absolute depth error is one percent, which is 5, and the depth difference between the depth map of the current frame and the depth map of the previous frame is 20, which is greater than 5, and the corresponding mark is the second environment change area .
  • the area formed by each pixel with a depth difference less than the preset absolute depth error is marked as the first environment change area, where the first area mask corresponding to the first environment change area is marked,
  • the corresponding area can be quickly identified according to the sub-area mask;
  • the area formed by each pixel with a depth difference greater than or equal to the preset absolute depth error is marked as the second environmental change area, where The second area mask corresponding to the second environment change area is marked to facilitate the subsequent smoothing process to quickly identify the corresponding area according to the sub-area mask.
  • Step 205 Obtain the first original depth value of the previous frame and the first original depth value of the current frame corresponding to each pixel in the first environment change area in the preset coordinate system, and compare the first similarity weight with the first original depth value of the previous frame. The product of an original depth value and the product of the third similarity weight and the first original depth value of the current frame are added together to obtain the first current frame depth value corresponding to each pixel in the first environment change area.
  • Step 206 Generate a second similarity weight corresponding to each pixel in the second environment change area according to the depth difference, the preset original smoothing coefficient after the reduction process, and the depth error value of the pixel point of the current frame.
  • Step 207 Obtain the second original depth value of the previous frame and the second original depth value of the current frame corresponding to each pixel in the second environment change area in the preset coordinate system, and compare the second similarity weight with the first frame of the previous frame. The product of the two original depth values and the product of the fourth similarity weight and the second original depth value of the current frame are added together to obtain the second current frame depth value corresponding to each pixel in the second environment change area.
  • the formula of w1 is s is the preset original smoothing coefficient.
  • the first area of environmental change is also the area of slow environmental change, which has relatively high smoothness, that is, relatively high credibility. This point belongs to the area of slow environmental change.
  • the enlargement process generates the first similarity weight.
  • the depth temporal consistency filtering based on depth change area detection focuses on the preprocessing of the depth map from the time dimension, which provides a time dimension for subsequent ToF depth map related applications such as gesture recognition, 3D modeling, somatosensory games, etc. Smoother and more stable depth data to achieve a better application experience.
  • the depth data filtering method of the embodiment of the present application obtains the depth difference between each pixel in two consecutive frames of depth maps; the depth difference is smaller than the preset absolute depth error.
  • the area is marked as the first environment change area; among them, the preset absolute depth error is determined according to the depth value of the depth map of each pixel in the previous frame and the preset error percentage; each pixel whose depth difference is greater than or equal to the preset absolute depth error.
  • the area formed by the dots is marked as the second environment change area; according to the depth difference, the preset original smoothing coefficient after the amplification process, and the pixel depth error value of the current frame, the first environment change area corresponding to each pixel in the first environment change area is generated.
  • the depth data filtering device includes: an acquisition module 501, a first marking module 502, a second marking module 503, and a first generation Module 504, second generation module 505 and processing module 506.
  • the first marking module 502 is used to mark the area formed by each pixel with a depth difference less than the preset absolute depth error as the first environment change area; wherein, according to the depth value of each pixel in the previous frame depth map and The preset error percentage determines the preset absolute depth error.
  • the second marking module 503 is configured to mark an area formed by pixels whose depth difference is greater than or equal to a preset absolute depth error as a second environment change area.
  • the first generating module 504 is configured to generate a first similarity weight corresponding to each pixel in the first environment change area according to the depth difference, the preset original smoothing coefficient after the amplification process, and the current frame pixel depth error value.
  • the second generation module 505 is configured to generate a second similarity weight corresponding to each pixel in the second environment change area according to the depth difference, the preset original smoothing coefficient after reduction processing, and the current frame pixel depth error value.
  • the processing module 506 is configured to perform filtering processing on the first environmental change area according to the first similarity weight, and perform filtering processing on the second environmental change area according to the second similarity weight.
  • the device further includes: a first mask processing module 507 and a second mask processing module 508, wherein,
  • the first mask processing module 507 is configured to mark the first area mask corresponding to the first environment change area.
  • the second mask processing module 508 is configured to mark the second area mask corresponding to the second environment change area.
  • the processing module 506 is specifically configured to obtain the first original depth value of the previous frame and the first original depth value of the current frame corresponding to each pixel in the first environment change area in the preset coordinate system.
  • the original depth value; the product of the first similarity weight and the first original depth value of the previous frame and the product of the third similarity weight and the first original depth value of the current frame are added together to obtain the first The first current frame depth value corresponding to each pixel in the environmental change area; wherein the sum of the first similarity weight and the third similarity weight is 1.
  • the processing module 506 is specifically configured to obtain the second original depth value of the previous frame and the second original depth value of the current frame corresponding to each pixel in the second environment change area in the preset coordinate system.
  • Original depth value adding the product of the second similarity weight and the second original depth value of the previous frame and the product of the fourth similarity weight and the second original depth value of the current frame to obtain the second The second current frame depth value corresponding to each pixel in the environment change area, wherein the sum of the second similarity weight and the fourth similarity weight is 1.
  • a preset formula is applied to generate a similarity weight based on the depth difference of each pixel, the preset original smoothing coefficient, and the current frame pixel depth error value.
  • the preset formula is: Wherein, s is a preset original smoothing coefficient, diff is the depth difference value, and ⁇ is the pixel depth error value of the current frame.
  • the depth data filtering device of the embodiment of the present application obtains the depth difference of each pixel point between two consecutive frames of depth maps; the depth difference is smaller than the preset absolute depth error.
  • the area is marked as the first environment change area; among them, the preset absolute depth error is determined according to the depth value of the depth map of each pixel in the previous frame and the preset error percentage; each pixel whose depth difference is greater than or equal to the preset absolute depth error.
  • the area formed by the dots is marked as the second environment change area; according to the depth difference, the preset original smoothing coefficient after the amplification process, and the pixel depth error value of the current frame, the first environment change area corresponding to each pixel in the first environment change area is generated.
  • Similarity weight according to the depth difference, the reduced preset original smoothing coefficient and the current frame pixel depth error value, the second similarity weight corresponding to each pixel in the second environmental change area is generated; according to the first similarity
  • the weight performs filtering processing on the first environmental change area
  • the second similarity weight performs filtering processing on the second environmental change area.
  • this application also proposes an electronic device, including a memory, a processor, and a computer program stored on the memory and capable of running on the processor.
  • the processor executes the computer program, it realizes the Filtering method of depth data.
  • the embodiment of the present application also proposes a computer-readable storage medium on which a computer program is stored.
  • the program is executed by a processor, the method for filtering depth data as described in the foregoing method embodiment is implemented.
  • first and second are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with “first” and “second” may explicitly or implicitly include at least one of the features. In the description of the present application, "a plurality of” means at least two, such as two, three, etc., unless specifically defined otherwise.
  • a "computer-readable medium” can be any device that can contain, store, communicate, propagate, or transmit a program for use by an instruction execution system, device, or device or in combination with these instruction execution systems, devices, or devices.
  • computer readable media include the following: electrical connections (electronic devices) with one or more wiring, portable computer disk cases (magnetic devices), random access memory (RAM), Read only memory (ROM), erasable and editable read only memory (EPROM or flash memory), fiber optic devices, and portable compact disk read only memory (CDROM).
  • the computer-readable medium may even be paper or other suitable media on which the program can be printed, because it can be used, for example, by optically scanning the paper or other media, and then editing, interpreting, or other suitable media if necessary. The program is processed in a manner to obtain the program electronically and then stored in the computer memory.
  • each part of this application can be implemented by hardware, software, firmware, or a combination thereof.
  • multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system.
  • Discrete logic gate circuits for implementing logic functions on data signals Logic circuit, application specific integrated circuit with suitable combinational logic gate, programmable gate array (PGA), field programmable gate array (FPGA), etc.
  • the functional units in the various embodiments of the present application may be integrated into one processing module, or each unit may exist alone physically, or two or more units may be integrated into one module.
  • the above-mentioned integrated modules can be implemented in the form of hardware or software functional modules. If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it may also be stored in a computer readable storage medium.
  • the aforementioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Remote Sensing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Radar, Positioning & Navigation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Electromagnetism (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

一种深度数据的滤波方法、装置、电子设备和可读存储介质,其中,方法包括:获取每个像素点在连续两帧深度图之间的深度差值;将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;分别对第一环境变化区域和第二环境变化区域进行不同的滤波处理。

Description

深度数据的滤波方法、装置、电子设备和可读存储介质
优先权信息
本申请请求2019年7月11日向中国国家知识产权局提交的、专利申请号为201910626650.X的专利申请的优先权和权益,并且通过参照将其全文并入此处。
技术领域
本申请涉及通信技术领域,尤其涉及一种深度数据的滤波方法、装置、电子设备和可读存储介质。
背景技术
通常,ToF(Time of Flight)传感器通过计算脉冲信号的飞行时间来确定传感器和物体之间的距离,由于测量过程中存在着各类不确定性,带来了多种误差,并且这些误差具有很大的随机性,造成了在测量范围内ToF的深度测量误差大约为1%。
在实际系统中,可以接受上述测量误差,但是希望传感器在有限的时间内可以达到时间一致性。
发明内容
本申请实施方式提供一种深度数据的滤波方法、装置、电子设备和可读存储介质。
本申请第一方面实施例提出一种深度数据的滤波方法,包括:
获取每个像素点在连续两帧深度图之间的深度差值;
将所述深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定所述预设绝对深度误差;
将所述深度差值大于等于所述预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;
根据所述深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与所述第一环境变化区域中各个像素点对应的第一相似度权重;
根据所述深度差值、缩小处理后的预设原始平滑系数和所述当前帧像素点深度误差值生成与所述第二环境变化区域中各个像素点对应的第二相似度权重;
根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
本申请第二方面实施例提出了一种深度数据的滤波装置,包括:
获取模块,用于获取每个像素点在连续两帧深度图之间的深度差值;
第一标记模块,用于将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定所述预设绝对深度误差;
第二标记模块,用于将所述深度差值大于等于所述预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;
第一生成模块,用于根据所述深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与所述第一环境变化区域中各个像素点对应的第一相似度权重;
第二生成模块,用于根据所述深度差值、缩小处理后的预设原始平滑系数和所述当前帧像素点深度误差值生成与所述第二环境变化区域中各个像素点对应的第二相似度权重;
处理模块,用于根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
本申请第三方面实施例提出了一种电子设备,图像传感器、存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,所述图像传感器与所述处理器电连接,所述处理器执行所述程序时,实现如前述方法实施例所述的深度数据的滤波方法。
本申请第四个方面实施例提出了一种计算机可读存储介质,其上存储有计算机程序,该程序被处理器执行时,实现如前述方法实施例所述的深度数据的滤波方法。
本申请附加的方面和优点将在下面的描述中部分给出,部分将从下面的描述中变得明显,或通过本申请的实践了解到。
附图说明
本申请上述的和/或附加的方面和优点从下面结合附图对实施例的描述中将变得明显和容易理解,其中:
图1为本申请实施例所提供的一种深度获取方法的流程示意图;
图2为本申请实施例所提供的一种深度数据的滤波方法的流程示意图;
图3为本申请实施例所提供的一种获取原始深度值的示意图;
图4为本申请实施例所提供的另一种深度数据的滤波方法的流程示意图;
图5是根据本申请一个实施例的一种深度数据的滤波装置的结构示意图;
图6是根据本申请一个实施例的另一种深度数据的滤波装置的结构示意图。
具体实施方式
下面详细描述本申请的实施例,所述实施例的示例在附图中示出,其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的,旨在用于解释本申请,而不能理解为对本申请的限制。
图2为本申请实施例所提供的一种深度数据的滤波方法的流程示意图。如图2所示,该方法包括以下步骤:
步骤101,获取每个像素点在连续两帧深度图之间的深度差值。
步骤102,将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域,其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差。
步骤103,将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域。
步骤104,根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
步骤105,根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
步骤106,根据第一相似度权重对第一环境变化区域进行滤波处理,以及第二相似度权重对第二环境变化区域进行滤波处理。
作为一种可能实现方式,获取第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值,将第一相似度权重与上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到第一环境变化区域中每个像素点对应的第一当前帧深度值;其中,第一相似度权重和第三相似度权重的和为1。
作为一种可能实现方式,获取第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值,将第二相似度权重与上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到第二环境变化区域中每个像素点对应的第二当前帧深度值,其中,第二相似度权重和第四相似度权重的和为1。
图4为本申请实施例所提供的另一种深度数据的滤波方法的流程示意图。如图4所示,该方法包括以下步骤:
步骤201,获取每个像素点在连续两帧深度图之间的深度差值。
步骤202,将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域,对第一环境变化区域标记对应的第一区域掩码。
步骤203,将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域,对第二环境变化区域标记对应的第二区域掩码。
步骤204,根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
步骤205,获取第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值,将第一相似度权重与上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到第一环境变化区域中每个像素点对应的第一当前帧深度值。
步骤206,根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
步骤207,获取第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值,将第二相似度权重与上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到第二环境变化区域中每个像素点对应的第二当前帧深度值。
其中,预设坐标系为笛卡尔坐标系,一个像素点当前帧深度图的深度值=上一帧深度图的深度*w1+当前帧深度图的原始深度*w2。
其中,w1的公式为
Figure PCTCN2020097464-appb-000001
s为预设原始平滑系数,diff为所述深度差值,表示该点在前后帧之间反射率差,σ为当前帧像素点深度误差值。
为了实现上述实施例,本申请还提出一种深度数据的滤波装置,如图5所示,深度数据的滤波装置包括:获取模块501、第一标记模块502、第二标记模块503、第一生成模块504、第二生成模块505和处理模块506。
其中,获取模块501,用于获取每个像素点在连续两帧深度图之间的深度差值。
第一标记模块502,用于将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差。
第二标记模块503,用于将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域。
第一生成模块504,用于根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
第二生成模块505,用于根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
处理模块506,用于根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
在本申请的一个实施例中,应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重。
在本申请的一个实施例中,预设公式为:
Figure PCTCN2020097464-appb-000002
其中,s为预设原始平滑系数,diff为所述深度差值,σ为所述当前帧像素点深度误差值。
为了实现上述实施例,本申请还提出一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,处理器执行计算机程序时,实现如前述实施例描述的深度数据的滤波方法。
为了实现上述实施例,本申请实施例还提出一种计算机可读存储介质,其上存储有计算机程序,该程序被处理器执行时实现如前述方法实施例所述的深度数据的滤波方法。
具体地,ToF传感器通过计算脉冲信号的飞行时间来确定传感器和物体之间的距离,比如
Figure PCTCN2020097464-appb-000003
其中d为深度,c为光速,t表示飞行时间;除以2是因为脉冲信号在传感器和物体之间内飞行了两次,基于上述背景技术的描述,可以了解到针对ToF深度数据的时间一致性滤波非常重要,ToF每一帧图像深度的获取方式如图1所示,ToF传感器发射经过调制的脉冲信号,待测量物体表面接收到脉冲信号并反射信号,然后ToF传感器接收到反射信号,并对多频相位图解码,接着根据标定参数对ToF数据进行误差修正,然后对多频信号去混叠,并将深度值由径向坐标系转换到笛卡尔坐标系,最后对深度值进行时间一致性滤波,输出时间维度上相对平滑的深度结果。
但是,上述方式是针对全画幅内的所有像素点进行滤波,会导致时间一致性滤波平滑性不足,造成深度数据在时间域抖动较大的技术问题,通过将深度图分为两个环境变化区域,并分区域选择不同的策略进行平滑处理,有效的使深度平缓变化区域在时间维度上深度值更为平滑,而深度快速变化区域又保持了原来的高动态性,具体如下:
下面参考附图描述本申请实施例的深度数据的滤波方法、装置、电子设备和可读存储介质。
图2为本申请实施例所提供的一种深度数据的滤波方法的流程示意图。如图2所示,该方法包括以下步骤:
步骤101,获取每个像素点在连续两帧深度图之间的深度差值。
步骤102,将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域,其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差。
步骤103,将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域。
具体地,可以获取每个像素点在连续两帧深度图之间比如在当前帧深度图与上一帧深度图之间的深度差值,并将该深度差值与预设绝对深度误差进行比较,从而确定该像素点需要标记为第一环境变化区域还是第二环境变化区域,即环境缓慢变化区域还是环境快速变化区域,其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差。
其中,预设绝对深度误差可以根据需要进行选择设置,比如一个像素点在前后帧深度值分别为[500,502],预设绝对深度误差为百分之一也就是5,该像素点在当前帧深度图与上一帧深度图之间的深度差值为2,也就是小于5,对应标记为第一环境变化区域;再比如一个像素点在前后帧深度值分别为[500,520],预设绝对深度误差为百分之一也就是5,该像素点在当前帧深度图与上一帧深度图之间的深度差值为20,也就是大于5,对应标记为第二环境变化区域。
由此,根据上述方式将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域,将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域。
步骤104,根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
步骤105,根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
步骤106,根据第一相似度权重对第一环境变化区域进行滤波处理,以及第二相似度权重对第二环境变化区域进行滤波处理。
因此,在确定好第一环境变化区域和第二环境变化区域后,需要分区域进行平滑处理,首先根据每个像素点的深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重和根据每个像素点的深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
其中,可以应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重,比如预设公式为:
Figure PCTCN2020097464-appb-000004
其中,s为预设原始平滑系数,diff为深度差值,σ为当前帧像素点深度误差值。
因此,第一环境变化区域也就是环境缓慢变化区域,具有比较高的平滑性,也就是可信度比较高,需要将预设原始平滑系数进行放大处理,从而提高第一相似度权重。
进一步地,可以通过上述公式根据每个像素点的深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
同理,第二环境变化区域也就是环境快速变化区域,具有比较低的平滑性,也就是可信度比较低,需要将预设原始平滑系数进行缩小处理,从而将第二相似度权重。
进一步地,也可以根据上述公式根据每个像素点的深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
最后,根据第一相似度权重对第一环境变化区域进行滤波处理,以及第二相似度权重对第二环境变化区域进行滤波处理。
具体地,首先需要获取当前帧深度图中每一个像素对应的深度值,具体地,如图3所示,首先ToF传感器采集原始相位图,单频模式下为四相位图,双频模式下为八相位图,接着由原始相位图计算每个像素点的I(相位余弦)Q(相位正弦)信号,并根据IQ信号计算每个像素点的相位和置信度,其中,置信度表示该像素点相位值的可信度,是该像素点能量大小的反应。
进一步地,根据ToF离线标定的内参在线修正几种误差,包括循环误差,温度误差,梯度误差,视差误差等,以及在双频去混叠前进行前滤波,分别过滤各频率模式下的噪声,进行双频去混叠,确定每个像素点的真实周期数,最后对去混叠结果进行后滤波,将深度值由径向坐标系转换到笛卡尔坐标系,也就是说上述预设坐标系优选笛卡尔坐标系。
其中,根据第一相似度权重对第一环境变化区域进行滤波处理的方式有很多种,比如直接根据第一相似度权重与像素点在相邻帧深度图对应的深度值进行处理,还是接着确定出第三相似度,结合第一相似度权重、第三相似度和像素点在相邻帧深度图对应的深度值进行处理,可以根据实际应用需要进行选择。
作为一种可能实现方式,获取第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值,将第一相似度权重与上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到第一 环境变化区域中每个像素点对应的第一当前帧深度值;其中,第一相似度权重和第三相似度权重的和为1。
需要说明的是,预设原始平滑系数是根据时间一致性滤波设置的原始经验值。
其中,根据第二相似度权重对第二环境变化区域进行滤波处理的方式有很多种,比如直接根据第二相似度权重与像素点在相邻帧深度图对应的深度值进行处理,还是接着确定出第四相似度,结合第二相似度权重、第四相似度和像素点在相邻帧深度图对应的深度值进行处理,可以根据实际应用需要进行选择。
作为一种可能实现方式,获取第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值,将第二相似度权重与上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到第二环境变化区域中每个像素点对应的第二当前帧深度值,其中,第二相似度权重和第四相似度权重的和为1。
综上,本申请实施例的深度数据的滤波方法,通过获取每个像素点在连续两帧深度图之间的深度差值;将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差;将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重;根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重;根据第一相似度权重对第一环境变化区域进行滤波处理,以及第二相似度权重对第二环境变化区域进行滤波处理。由此,有效解决了现有技术中导致时间一致性滤波平滑性不足,造成深度数据在时间域抖动较大的技术问题,通过将深度图分为两个环境变化区域,并分区域选择不同的策略进行平滑处理,有效的使深度平缓变化区域在时间维度上深度值更为平滑,而深度快速变化区域又保持了原来的高动态性。
图4为本申请实施例所提供的另一种深度数据的滤波方法的流程示意图。如图4所示,该方法包括以下步骤:
步骤201,获取每个像素点在连续两帧深度图之间的深度差值。
步骤202,将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域,对第一环境变化区域标记对应的第一区域掩码。
步骤203,将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域,对第二环境变化区域标记对应的第二区域掩码。
具体地,可以获取每个像素点在连续两帧深度图之间比如在当前帧深度图与上一帧深度图之间的深度差值,并将该深度差值与预设绝对深度误差进行比较,从而确定该像素点需要标记为第一环境变化区域还是第二环境变化区域,即环境缓慢变化区域还是环境快速变化区域。
其中,预设绝对深度误差可以根据需要进行选择设置,比如一个像素点在前后帧深度值分别为[500,502],预设绝对深度误差为百分之一也就是5,该像素点在当前帧深度图与上一帧深度图之间的深度差值为2,也就是小于5,对应标记为第一环境变化区域;再比如一个像素点在前后帧深度值分别为[500,520],预设绝对深度误差为百分之一也就是5,该像素点在当前帧深度图与上一帧深度图之间的深度差值为20,也就是大于5,对应标记为第二环境变化区域。
由此,根据上述方式将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域,其中,通过对第一环境变化区域标记对应的第一区域掩码,以方便后续进行平滑处理时能够根据分区域掩码快速识别对应的区域;将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区,其中,通过对第二环境变化区域标记对应的第二区域掩码,以方便后续进行平滑处理时能够根据分区域掩码快速识别对应的区域。
步骤204,根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
步骤205,获取第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值,将第一相似度权重与上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到第一环境变化区域中每个像素点对应的第一当前帧深度值。
步骤206,根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
步骤207,获取第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值,将第二相似度权重与上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到第二环境变化区域中每个像素点对应的第二当前帧深度值。
其中,预设坐标系为笛卡尔坐标系,一个像素点当前帧深度图的深度值=上一帧深度图的深度*w1+当前帧深度图的原始深度*w2。
其中,w1的公式为
Figure PCTCN2020097464-appb-000005
s为预设原始平滑系数,第一环境变化区域 也就是环境缓慢变化区域,具有比较高的平滑性,也就是可信度比较高,该点属于环境缓慢变化区域,对预设原始平滑系数进行放大处理生成第一相似度权重。
另外,第二环境变化区域也就是环境快速变化区域,具有比较低的平滑性,也就是可信度比较低,该点属于环境快速变化区域,对预设原始平滑系数进行缩小处理生成第二相似度权重;diff为所述深度差值,表示该点在前后帧之间反射率差。
需要说明的是,σ为当前帧像素点深度误差值,σ=dep*1%,dep为当前帧深度图的原始深度,在两帧之间深度值分布满足正态分布,则认为时间域上噪声较小,即正态分布的σ很小,即具有较高平滑性,否则为低平滑性。
由此,基于深度变化区域检测的深度时间一致性滤波,着重从时间维度上对深度图进行预处理,为后续ToF深度图相关应用如手势识别,三维建模,体感游戏等提供了时间维度上更为平滑稳定的深度数据,实现更好的应用体验。
综上,本申请实施例的深度数据的滤波方法,通过获取每个像素点在连续两帧深度图之间的深度差值;将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差;将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重;根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重;根据第一相似度权重对第一环境变化区域进行滤波处理,以及第二相似度权重对第二环境变化区域进行滤波处理。由此,有效解决了现有技术中导致时间一致性滤波平滑性不足,造成深度数据在时间域抖动较大的技术问题,通过将深度图分为两个环境变化区域,并分区域选择不同的策略进行平滑处理,有效的使深度平缓变化区域在时间维度上深度值更为平滑,而深度快速变化区域又保持了原来的高动态性。
为了实现上述实施例,本申请还提出一种深度数据的滤波装置,如图5所示,深度数据的滤波装置包括:获取模块501、第一标记模块502、第二标记模块503、第一生成模块504、第二生成模块505和处理模块506。
其中,获取模块501,用于获取每个像素点在连续两帧深度图之间的深度差值。
第一标记模块502,用于将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定预设绝对深度误差。
第二标记模块503,用于将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域。
第一生成模块504,用于根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重。
第二生成模块505,用于根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重。
处理模块506,用于根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
在本申请的一个实施例中,如图6所示,在如图5所示的基础上,该装置还包括:第一掩码处理模块507和第二掩码处理模块508,其中,
第一掩码处理模块507,用于对所述第一环境变化区域标记对应的第一区域掩码。
第二掩码处理模块508,用于对所述第二环境变化区域标记对应的第二区域掩码。
在本申请的一个实施例中,处理模块506,具体用于获取所述第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值;将所述第一相似度权重与所述上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到所述第一环境变化区域中每个像素点对应的第一当前帧深度值;其中,所述第一相似度权重和所述第三相似度权重的和为1
在本申请的一个实施例中,处理模块506,具体用于获取所述第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值;将所述第二相似度权重与所述上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到所述第二环境变化区域中每个像素点对应的第二当前帧深度值,其中,所述第二相似度权重和所述第四相似度权重的和为1。
在本申请的一个实施例中,应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重。
在本申请的一个实施例中,预设公式为:
Figure PCTCN2020097464-appb-000006
其中,s为预设原始平滑系数,diff为所述深度差值,σ为所述当前帧像素点深度误差值。
需要说明的是,前述集中在深度数据的滤波方法实施例中的解释说明,也适用于本申请实施例的深度数据的滤波装置,在此不再对其实施细节和技术效果赘述。
综上,本申请实施例的深度数据的滤波装置,通过获取每个像素点在连续两帧深度图之间的深度差值;将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定 预设绝对深度误差;将深度差值大于等于预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;根据深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第一环境变化区域中各个像素点对应的第一相似度权重;根据深度差值、缩小处理后的预设原始平滑系数和当前帧像素点深度误差值生成与第二环境变化区域中各个像素点对应的第二相似度权重;根据第一相似度权重对第一环境变化区域进行滤波处理,以及第二相似度权重对第二环境变化区域进行滤波处理。由此,有效解决了现有技术中导致时间一致性滤波平滑性不足,造成深度数据在时间域抖动较大的技术问题,通过将深度图分为两个环境变化区域,并分区域选择不同的策略进行平滑处理,有效的使深度平缓变化区域在时间维度上深度值更为平滑,而深度快速变化区域又保持了原来的高动态性。
为了实现上述实施例,本申请还提出一种电子设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,处理器执行计算机程序时,实现如前述实施例描述的深度数据的滤波方法。
为了实现上述实施例,本申请实施例还提出一种计算机可读存储介质,其上存储有计算机程序,该程序被处理器执行时实现如前述方法实施例所述的深度数据的滤波方法。
在本说明书的描述中,参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本申请的至少一个实施例或示例中。在本说明书中,对上述术语的示意性表述不必须针对的是相同的实施例或示例。而且,描述的具体特征、结构、材料或者特点可以在任一个或多个实施例或示例中以合适的方式结合。此外,在不相互矛盾的情况下,本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。
此外,术语“第一”、“第二”仅用于描述目的,而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此,限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。在本申请的描述中,“多个”的含义是至少两个,例如两个,三个等,除非另有明确具体的限定。
流程图中或在此以其他方式描述的任何过程或方法描述可以被理解为,表示包括一个或更多个用于实现定制逻辑功能或过程的步骤的可执行指令的代码的模块、片段或部分,并且本申请的优选实施方式的范围包括另外的实现,其中可以不按所示出或讨论的顺序,包括根据所涉及的功能按基本同时的方式或按相反的顺序,来执行功能,这应被本申请的实施例所属技术领域的技术人员所理解。
在流程图中表示或在此以其他方式描述的逻辑和/或步骤,例如,可以被认为是用于实现逻辑功能的可执行指令的定序列表,可以具体实现在任何计算机可读介质中,以供指令执行系统、装置或设备(如基于计算机的系统、包括处理器的系统或其他可以从指令执行系统、装置或设备取指令并执行指令的系统)使用,或结合这些指令执行系统、装置或设备而使用。就本说明书而言,"计算机可读介质"可以是任何可以包含、存储、通信、传播或传输程序以供指令执行系统、装置或设备或结合这些指令执行系统、装置或设备而使用的装置。计算机可读介质的更具体的示例(非穷尽性列表)包括以下:具有一个或多个布线的电连接部(电子装置),便携式计算机盘盒(磁装置),随机存取存储器(RAM),只读存储器(ROM),可擦除可编辑只读存储器(EPROM或闪速存储器),光纤装置,以及便携式光盘只读存储器(CDROM)。另外,计算机可读介质甚至可以是可在其上打印所述程序的纸或其他合适的介质,因为可以例如通过对纸或其他介质进行光学扫描,接着进行编辑、解译或必要时以其他合适方式进行处理来以电子方式获得所述程序,然后将其存储在计算机存储器中。
应当理解,本申请的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式中,多个步骤或方法可以用存储在存储器中且由合适的指令执行系统执行的软件或固件来实现。如,如果用硬件来实现和在另一实施方式中一样,可用本领域公知的下列技术中的任一项或他们的组合来实现:具有用于对数据信号实现逻辑功能的逻辑门电路的离散逻辑电路,具有合适的组合逻辑门电路的专用集成电路,可编程门阵列(PGA),现场可编程门阵列(FPGA)等。
本技术领域的普通技术人员可以理解实现上述实施例方法携带的全部或部分步骤是可以通过程序来指令相关的硬件完成,所述的程序可以存储于一种计算机可读存储介质中,该程序在执行时,包括方法实施例的步骤之一或其组合。
此外,在本申请各个实施例中的各功能单元可以集成在一个处理模块中,也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个模块中。上述集成的模块既可以采用硬件的形式实现,也可以采用软件功能模块的形式实现。所述集成的模块如果以软件功能模块的形式实现并作为独立的产品销售或使用时,也可以存储在一个计算机可读取存储介质中。
上述提到的存储介质可以是只读存储器,磁盘或光盘等。尽管上面已经示出和描述了本申请的实施例,可以理解的是,上述实施例是示例性的,不能理解为对本申请的限制,本领域的普通技术人员在本申请的范围内可以对上述实施例进行变化、修改、替换和变型。

Claims (20)

  1. 一种深度数据的滤波方法,其特征在于,包括以下步骤:
    获取每个像素点在连续两帧深度图之间的深度差值;
    将所述深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定所述预设绝对深度误差;
    将所述深度差值大于等于所述预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;
    根据所述深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与所述第一环境变化区域中各个像素点对应的第一相似度权重;
    根据所述深度差值、缩小处理后的预设原始平滑系数和所述当前帧像素点深度误差值生成与所述第二环境变化区域中各个像素点对应的第二相似度权重;
    根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
  2. 如权利要求1所述的方法,其特征在于,还包括:
    对所述第一环境变化区域标记对应的第一区域掩码。
  3. 如权利要求1所述的方法,其特征在于,还包括:
    对所述第二环境变化区域标记对应的第二区域掩码。
  4. 如权利要求1所述的方法,其特征在于,所述根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,包括:
    获取所述第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值;
    将所述第一相似度权重与所述上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到所述第一环境变化区域中每个像素点对应的第一当前帧深度值;其中,所述第一相似度权重和所述第三相似度权重的和为1。
  5. 如权利要求1所述的方法,其特征在于,所述根据所述第二相似度权重对所述第二环境变化区域进行滤波处理,包括:
    获取所述第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值;
    将所述第二相似度权重与所述上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到所述第二环境变化区域中每个像素点对应的第二当前帧深度值,其中,所述第二相似度权重和所述第四相似度权重的和为1。
  6. 如权利要求4或5所述的方法,其特征在于,
    应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重;
    所述预设公式为:
    Figure PCTCN2020097464-appb-100001
    其中,s为所述预设原始平滑系数,diff为所述深度差值,σ为所述当前帧像素点深度误差值。
  7. 一种深度数据的滤波装置,其特征在于,包括:
    获取模块,用于获取每个像素点在连续两帧深度图之间的深度差值;
    第一标记模块,用于将深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定所述预设绝对深度误差;
    第二标记模块,用于将所述深度差值大于等于所述预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;
    第一生成模块,用于根据所述深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与所述第一环境变化区域中各个像素点对应的第一相似度权重;
    第二生成模块,用于根据所述深度差值、缩小处理后的预设原始平滑系数和所述当前帧像素点深度误差值生成与所述第二环境变化区域中各个像素点对应的第二相似度权重;
    处理模块,用于根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
  8. 如权利要求7所述的装置,其特征在于,
    应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重;
    所述预设公式为:
    Figure PCTCN2020097464-appb-100002
    其中,s为所述预设原始平滑系数,diff为所述深度差值,σ为所述当前帧像素点深度误差值。
  9. 一种电子设备,其特征在于,包括:图像传感器、存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,所述图像传感器与所述处理器电连接,所述处理器执行所述程序时,实现以下深度数据的滤波方法:
    获取每个像素点在连续两帧深度图之间的深度差值;
    将所述深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定所述预设绝对深度误差;
    将所述深度差值大于等于所述预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;
    根据所述深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与所述第一环境变化区域中各个像素点对应的第一相似度权重;
    根据所述深度差值、缩小处理后的预设原始平滑系数和所述当前帧像素点深度误差值生成与所述第二环境变化区域中各个像素点对应的第二相似度权重;
    根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
  10. 如权利要求9所述的电子设备,其特征在于,所述处理器执行所述程序时还实现:
    对所述第一环境变化区域标记对应的第一区域掩码。
  11. 如权利要求9所述的电子设备,其特征在于,所述处理器执行所述程序时还实现:
    对所述第二环境变化区域标记对应的第二区域掩码。
  12. 如权利要求9所述的电子设备,其特征在于,所述处理器执行所述程序时实现:
    获取所述第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值;
    将所述第一相似度权重与所述上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到所述第一环境变化区域中每个像素点对应的第一当前帧深度值;其中,所述第一相似度权重和所述第三相似度权重的和为1。
  13. 如权利要求9所述的电子设备,其特征在于,所述处理器执行所述程序时实现:
    获取所述第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值;
    将所述第二相似度权重与所述上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到所述第二环境变化区域中每个像素点对应的第二当前帧深度值,其中,所述第二相似度权重和所述第四相似度权重的和为1。
  14. 如权利要求12或13所述的电子设备,其特征在于,所述处理器执行所述程序时实现:
    应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重;
    所述预设公式为:
    Figure PCTCN2020097464-appb-100003
    其中,s为所述预设原始平滑系数,diff为所述深度差值,σ为所述当前帧像素点深度误差值。
  15. 一种计算机可读存储介质,其上存储有计算机程序,其特征在于,所述计算机程序被处理器执行时实现以下深度数据的滤波方法:
    获取每个像素点在连续两帧深度图之间的深度差值;
    将所述深度差值小于预设绝对深度误差的各个像素点所构成的区域标记为第一环境变化区域;其中,根据每个像素点在前帧深度图的深度值和预设误差百分比确定所述预设绝对深度误差;
    将所述深度差值大于等于所述预设绝对深度误差的各个像素点所构成的区域标记为第二环境变化区域;
    根据所述深度差值、放大处理后的预设原始平滑系数和当前帧像素点深度误差值生成与所述第一环境变化区域中各个像素点对应的第一相似度权重;
    根据所述深度差值、缩小处理后的预设原始平滑系数和所述当前帧像素点深度误差值生成与所述第二环境变化区域中各个像素点对应的第二相似度权重;
    根据所述第一相似度权重对所述第一环境变化区域进行滤波处理,以及所述第二相似度权重对所述第二环境变化区域进行滤波处理。
  16. 如权利要求15所述的计算机可读存储介质,其特征在于,所述计算机程序被处理器执行时还实现:
    对所述第一环境变化区域标记对应的第一区域掩码。
  17. 如权利要求15所述的计算机可读存储介质,其特征在于,所述计算机程序被处理器执行时还实现:
    对所述第二环境变化区域标记对应的第二区域掩码。
  18. 如权利要求15所述的计算机可读存储介质,其特征在于,所述计算机程序被处理器执行时实现:
    获取所述第一环境变化区域中每个像素点在预设坐标系下对应的上一帧第一原始深度值和当前帧第一原始深度值;
    将所述第一相似度权重与所述上一帧第一原始深度值的乘积和第三相似度权重与当前帧第一原始深度值的乘积进行相加处理得到所述第一环境变化区域中每个像素点对应的第一当前帧深度值;其中,所述第一相似度权重和所述第三相似度权重的和为1。
  19. 如权利要求15所述的计算机可读存储介质,其特征在于,所述计算机程序被处理器执行时实现:
    获取所述第二环境变化区域中每个像素点在预设坐标系下对应的上一帧第二原始深度值和当前帧第二原始深度值;
    将所述第二相似度权重与所述上一帧第二原始深度值的乘积和第四相似度权重与当前帧第二原始深度值的乘积进行相加处理得到所述第二环境变化区域中每个像素点对应的第二当前帧深度值,其中,所述第二相似度权重和所述第四相似度权重的和为1。
  20. 如权利要求18或19所述的计算机可读存储介质,其特征在于,所述计算机程序被处理器执行时实现:
    应用预设公式根据每个像素点的深度差值、预设原始平滑系数和当前帧像素点深度误差值生成相似度权重;
    所述预设公式为:
    Figure PCTCN2020097464-appb-100004
    其中,s为所述预设原始平滑系数,diff为所述深度差值,σ为所述当前帧像素点深度误差值。
PCT/CN2020/097464 2019-07-11 2020-06-22 深度数据的滤波方法、装置、电子设备和可读存储介质 WO2021004261A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP20837151.8A EP3975107A4 (en) 2019-07-11 2020-06-22 DEPTH DATA FILTERING METHOD AND APPARATUS, ELECTRONIC DEVICE AND READABLE STORAGE MEDIA
US17/559,103 US20220114744A1 (en) 2019-07-11 2021-12-22 Depth data filtering method and apparatus, electronic device, and readable storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910626650.X 2019-07-11
CN201910626650.XA CN110400273B (zh) 2019-07-11 2019-07-11 深度数据的滤波方法、装置、电子设备和可读存储介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/559,103 Continuation US20220114744A1 (en) 2019-07-11 2021-12-22 Depth data filtering method and apparatus, electronic device, and readable storage medium

Publications (1)

Publication Number Publication Date
WO2021004261A1 true WO2021004261A1 (zh) 2021-01-14

Family

ID=68324718

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/097464 WO2021004261A1 (zh) 2019-07-11 2020-06-22 深度数据的滤波方法、装置、电子设备和可读存储介质

Country Status (4)

Country Link
US (1) US20220114744A1 (zh)
EP (1) EP3975107A4 (zh)
CN (1) CN110400273B (zh)
WO (1) WO2021004261A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114911786A (zh) * 2022-05-30 2022-08-16 上海华兴数字科技有限公司 钻孔深度曲线优化方法、装置、设备及作业机械

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110400273B (zh) * 2019-07-11 2022-03-22 Oppo广东移动通信有限公司 深度数据的滤波方法、装置、电子设备和可读存储介质
CN111080561A (zh) * 2019-11-08 2020-04-28 上海航天控制技术研究所 一种时域高通滤波方法
CN112712477A (zh) * 2020-12-21 2021-04-27 东莞埃科思科技有限公司 结构光模组的深度图像评价方法及其装置
CN112907480B (zh) * 2021-03-11 2023-05-09 北京格灵深瞳信息技术股份有限公司 点云表面波纹去除方法、装置、终端及存储介质
US11941787B2 (en) * 2021-08-23 2024-03-26 Microsoft Technology Licensing, Llc Denoising depth data of low-signal pixels

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103413276A (zh) * 2013-08-07 2013-11-27 清华大学深圳研究生院 一种基于纹理分布特征的深度增强方法
CN104394399A (zh) * 2014-10-31 2015-03-04 天津大学 深度视频编码的三边滤波方法
CN104683783A (zh) * 2015-01-08 2015-06-03 电子科技大学 一种自适应深度图滤波方法
EP3385903A1 (en) * 2017-04-05 2018-10-10 Samsung Electronics Co., Ltd. Method and apparatus for refining depth image
CN110400273A (zh) * 2019-07-11 2019-11-01 Oppo广东移动通信有限公司 深度数据的滤波方法、装置、电子设备和可读存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9014462B2 (en) * 2010-11-10 2015-04-21 Panasonic Intellectual Property Management Co., Ltd. Depth information generating device, depth information generating method, and stereo image converter
KR20120057216A (ko) * 2010-11-26 2012-06-05 삼성전자주식회사 깊이 센서, 상기 깊이 센서의 노이즈 감소 방법, 및 상기 깊이 센서를 포함하는 신호 처리 시스템
KR101938205B1 (ko) * 2012-03-19 2019-01-14 한국전자통신연구원 깊이 영상 필터링 방법 및 그 장치
CN105335947B (zh) * 2014-05-26 2019-03-01 富士通株式会社 图像去噪方法和图像去噪装置
AU2014218390B2 (en) * 2014-08-27 2017-01-12 Canon Kabushiki Kaisha Method, system and apparatus for forming a high resolution depth map
CN105139401A (zh) * 2015-08-31 2015-12-09 山东中金融仕文化科技股份有限公司 一种深度图中深度的可信度的评估方法
CN107958446B (zh) * 2016-10-17 2023-04-07 索尼公司 信息处理设备和信息处理方法
CN109191506B (zh) * 2018-08-06 2021-01-29 深圳看到科技有限公司 深度图的处理方法、系统及计算机可读存储介质
CN109712192B (zh) * 2018-11-30 2021-03-23 Oppo广东移动通信有限公司 摄像模组标定方法、装置、电子设备及计算机可读存储介质
CN109636748B (zh) * 2018-12-04 2021-03-16 青岛小鸟看看科技有限公司 一种深度图像的去噪方法、装置和电子设备

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103413276A (zh) * 2013-08-07 2013-11-27 清华大学深圳研究生院 一种基于纹理分布特征的深度增强方法
CN104394399A (zh) * 2014-10-31 2015-03-04 天津大学 深度视频编码的三边滤波方法
CN104683783A (zh) * 2015-01-08 2015-06-03 电子科技大学 一种自适应深度图滤波方法
EP3385903A1 (en) * 2017-04-05 2018-10-10 Samsung Electronics Co., Ltd. Method and apparatus for refining depth image
CN110400273A (zh) * 2019-07-11 2019-11-01 Oppo广东移动通信有限公司 深度数据的滤波方法、装置、电子设备和可读存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3975107A4 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114911786A (zh) * 2022-05-30 2022-08-16 上海华兴数字科技有限公司 钻孔深度曲线优化方法、装置、设备及作业机械

Also Published As

Publication number Publication date
EP3975107A1 (en) 2022-03-30
CN110400273A (zh) 2019-11-01
US20220114744A1 (en) 2022-04-14
EP3975107A4 (en) 2022-08-17
CN110400273B (zh) 2022-03-22

Similar Documents

Publication Publication Date Title
WO2021004261A1 (zh) 深度数据的滤波方法、装置、电子设备和可读存储介质
WO2021004262A1 (zh) 深度图处理方法及装置、电子设备和可读存储介质
WO2021004260A1 (zh) 深度图处理方法和装置
US20220084225A1 (en) Depth Map Processing Method, Electronic Device and Readable Storage Medium
Liu et al. Rapid and automatic 3D body measurement system based on a GPU–Steger line detector
CN110838122A (zh) 点云的分割方法、装置及计算机存储介质
WO2021115040A1 (zh) 图像校正方法、装置、终端设备和存储介质
JP2015156136A (ja) 算出装置、方法及びプログラム
US11961246B2 (en) Depth image processing method and apparatus, electronic device, and readable storage medium
CN110390690A (zh) 深度图处理方法和装置
CN111161153B (zh) 宽景拼接方法、装置及存储介质
CN104361554A (zh) 一种基于血管内超声图像的血管外膜自动检测方法
CN111145264B (zh) 多传感器的标定方法、装置及计算设备
Zhao et al. Comparison of the existing tool localisation methods on two‐dimensional ultrasound images and their tracking results
CN113643311A (zh) 一种对边界误差鲁棒的图像分割方法和装置
WO2010089938A1 (ja) 回転推定装置、回転推定方法および記録媒体
JP2020034525A (ja) 情報処理装置及びその方法及びコンピュータプログラム
CN110390656B (zh) 深度数据的滤波方法、装置、电子设备和可读存储介质
CN110400272B (zh) 深度数据的滤波方法、装置、电子设备和可读存储介质
JP7152506B2 (ja) 撮像装置
CN110415287B (zh) 深度图的滤波方法、装置、电子设备和可读存储介质
CN111833397B (zh) 一种面向测向目标定位的数据转换方法和装置
WO2023061355A1 (zh) 速度检测方法、装置、设备及可读存储介质
CN115115700B (zh) 物体的姿态估计方法、装置、电子设备及存储介质
JP2023121453A (ja) 情報処理システム、推論方法および推論プログラム

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: 20837151

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2020837151

Country of ref document: EP

Effective date: 20211224