CN110782391A - Image processing method and device in driving simulation scene and storage medium - Google Patents
Image processing method and device in driving simulation scene and storage medium Download PDFInfo
- Publication number
- CN110782391A CN110782391A CN201910850744.5A CN201910850744A CN110782391A CN 110782391 A CN110782391 A CN 110782391A CN 201910850744 A CN201910850744 A CN 201910850744A CN 110782391 A CN110782391 A CN 110782391A
- Authority
- CN
- China
- Prior art keywords
- noise
- value
- pixel
- pixel point
- image
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 181
- 238000003672 processing method Methods 0.000 title claims abstract description 26
- 238000012545 processing Methods 0.000 claims abstract description 45
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000009877 rendering Methods 0.000 claims abstract description 18
- 230000015654 memory Effects 0.000 description 26
- 230000006870 function Effects 0.000 description 12
- 238000004364 calculation method Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010276 construction Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000012360 testing method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 206010034960 Photophobia Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 208000013469 light sensitivity Diseases 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/04—Context-preserving transformations, e.g. by using an importance map
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention provides an image processing method, an image processing device and a storage medium in a driving simulation scene; the method comprises the following steps: acquiring a simulation image output by an image rendering engine of a vehicle; determining the pixel value of each pixel point in the simulation image; generating a random number corresponding to each pixel value based on the pixel value of each pixel point in the simulation image; based on the pixel value of each pixel point and the corresponding random number, searching a noise model to obtain a Poisson noise value corresponding to each pixel point, wherein the noise value in the noise model obeys Poisson distribution; and generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point. By the method and the device, the generation efficiency of the noise image in the driving simulation scene can be improved.
Description
Technical Field
The invention relates to the technical field of automatic driving, in particular to an image processing method and device in a driving simulation scene and a storage medium.
Background
An automatic driving vehicle is also called an unmanned vehicle or a computer driving vehicle, and is an intelligent vehicle which can realize that the vehicle automatically travels along a road in an unmanned state. Generally, before an autonomous vehicle travels on an actual traffic road, a lot of tests are required to check the safety and stability of the autonomous vehicle.
In consideration of road safety, a driving simulation platform is provided in the related art to construct a driving anti-simulation scene so as to perform simulation test on the automatic driving vehicle. However, in the related art, when noise simulation is performed on an image in a driving simulation scene, a method of generating a poisson distribution random variable is used to obtain a noise value, and then image processing is performed, so that the calculation amount involved in the whole operation process is huge, the processing time is long, and the generation efficiency of a noise image is reduced.
Disclosure of Invention
The embodiment of the invention provides an image processing method and device in a driving simulation scene and a storage medium, which can improve the generation efficiency of a noise image in the driving simulation scene.
The technical scheme of the embodiment of the invention is realized as follows:
the embodiment of the invention provides an image processing method in a driving simulation scene, which comprises the following steps:
acquiring a simulation image output by an image rendering engine of a vehicle;
determining the pixel value of each pixel point in the simulation image;
generating a random number corresponding to each pixel value based on the pixel value of each pixel point in the simulation image;
based on the pixel value of each pixel point and the corresponding random number, searching a noise model to obtain a Poisson noise value corresponding to each pixel point, wherein the noise value in the noise model obeys Poisson distribution;
and generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point.
An embodiment of the present invention further provides an image processing apparatus in a driving simulation scene, including:
the system comprises an acquisition unit, a display unit and a control unit, wherein the acquisition unit is used for acquiring a simulation image output by an image rendering engine of a vehicle;
the determining unit is used for determining the pixel value of each pixel point in the simulation image;
the first generation unit is used for generating random numbers corresponding to pixel values of all pixels in the simulation image based on the pixel values of all the pixels in the simulation image;
the searching unit is used for searching and obtaining a Poisson noise value corresponding to each pixel point from a noise model based on the pixel value of each pixel point and the corresponding random number, and the noise value in the noise model obeys Poisson distribution;
and the second generation unit is used for generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point.
In the above scheme, the apparatus further comprises:
the model building unit is used for determining the gray value range and the noise level corresponding to the noise model;
determining the noise amplitude corresponding to each gray value belonging to the gray value range based on the gray value range and the noise level;
and generating the noise model based on the noise amplitude corresponding to each gray value.
In the above scheme, the model construction unit is configured to determine each gray value in the gray value range;
and determining the noise amplitude corresponding to each gray value based on the inverse proportional relation between the preset gray value and the noise amplitude.
In the foregoing scheme, the model building unit is configured to determine a noise value set corresponding to each gray value based on the noise amplitude corresponding to each gray value and a poisson distribution relationship built based on the noise amplitude;
respectively acquiring a random number set corresponding to each gray value;
respectively determining Poisson noise values corresponding to all random numbers in all random number sets based on the noise value sets and the random number sets corresponding to all gray values;
and generating the noise model based on the Poisson noise value corresponding to each random number in each random number set.
In the above scheme, the searching unit is configured to determine a gray value corresponding to the pixel point based on pixel values of the pixel point on red R, green G, and blue B color channels;
and searching to obtain the Poisson noise value corresponding to each pixel point from the noise model based on the determined gray value and the random number corresponding to the pixel point.
In the above scheme, the second generating unit is configured to determine the sum of the pixel value of each pixel point and the corresponding poisson noise value, respectively, to obtain a first target pixel value corresponding to each pixel point;
and generating a noise image for driving simulation based on the first target pixel value of each pixel point.
In the above scheme, the searching unit is configured to determine the gaussian noise value corresponding to each pixel point based on the random number corresponding to each pixel point.
In the above scheme, the second generating unit is configured to determine the sum of the pixel value of each pixel point and the corresponding noise value and gaussian noise value, respectively, to obtain a second target pixel value corresponding to each pixel point;
and generating a noise image for driving simulation based on the second target pixel value of each pixel point.
In the above scheme, the searching unit is configured to search, by using the graphics processor, from the noise model based on the pixel value of each pixel point and the corresponding random number, to obtain the poisson noise value corresponding to each pixel point.
An embodiment of the present invention further provides an image processing apparatus in a driving simulation scene, including:
a memory for storing executable instructions;
and the processor is used for realizing the image processing method in the driving simulation scene provided by the embodiment of the invention when the executable instructions stored in the memory are executed.
The embodiment of the invention also provides a storage medium, which stores executable instructions and is used for causing a processor to execute the executable instructions so as to realize the image processing method in the driving simulation scene provided by the embodiment of the invention.
The embodiment of the invention has the following beneficial effects:
generating a random number corresponding to each pixel value by acquiring the pixel value of each pixel point in the simulation image, searching a Poisson noise value corresponding to each pixel point in a preset noise model based on the pixel value and the corresponding random number, and adding the Poisson noise value and the pixel value of each corresponding pixel point to obtain a noise image for driving simulation; because the noise model is pre-constructed aiming at the pixel value of each pixel point, the Poisson noise value corresponding to each pixel value can be directly searched in the noise model when the image processing is carried out so as to realize the noise adding processing of the simulation image, thereby reducing the calculation amount and improving the generation efficiency of the noise image.
Drawings
FIG. 1 is a schematic diagram of an architecture of an image processing system in a driving simulation scenario provided by an embodiment of the present invention;
FIG. 2 is a schematic structural diagram of an image processing apparatus in a driving simulation scenario according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating an image processing method in a driving simulation scenario according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart of noise model construction in a driving simulation scenario according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart of adding Poisson noise to a simulation image according to an embodiment of the present invention;
FIG. 6 is a schematic flow chart of adding Gaussian noise to a simulated image according to an embodiment of the present invention;
FIG. 7 is a flowchart illustrating an image processing method in a driving simulation scenario according to an embodiment of the present invention;
fig. 8 is a schematic flow chart of noise simulation implemented based on a poisson dictionary according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of an image processing apparatus in a driving simulation scene according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail with reference to the accompanying drawings, the described embodiments should not be construed as limiting the present invention, and all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict.
In the description that follows, references to the terms "first", "second", and the like, are intended only to distinguish similar objects and not to indicate a particular ordering for the objects, it being understood that "first", "second", and the like may be interchanged under certain circumstances or sequences of events to enable embodiments of the invention described herein to be practiced in other than the order illustrated or described herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to be limiting of the invention.
Before further detailed description of the embodiments of the present invention, terms and expressions mentioned in the embodiments of the present invention are explained, and the terms and expressions mentioned in the embodiments of the present invention are applied to the following explanations.
1) The automatic driving is to guide and decide a vehicle driving task without testing the physical driving operation of a driver, and replace the control behavior of the driver to ensure that the vehicle can complete the function of safe driving.
2) The automatic driving simulation scene is a virtual traffic scene which is manufactured by a Unity3D development component and used for simulating automatic driving, and is mainly divided into four types, namely a natural driving scene, a dangerous working condition scene, a legal standard scene and a parameter recombination scene.
3) The pixel value, the value assigned by the computer when the image is digitized, represents the average luminance information for a small square in the image.
4) The gray scale value, representing the color depth of the dots in a black and white image, or the shade of the color of the image, typically ranges from 0 to 255.
5) Poisson noise, which refers to a type of noise whose probability density function obeys a poisson distribution, is generally generated by the fluctuation of the gray value of a pixel due to the intensity of a light source.
6) Gaussian noise refers to a type of noise whose probability density function follows a gaussian distribution.
7) Graphics Processing Unit (GPU), a microprocessor dedicated to image computing work on personal computers, workstations, gaming machines, and some mobile devices.
Before further describing the embodiments of the present invention in detail, the related art related to the embodiments of the present invention will be described first.
Image noise is generated during image capturing and cannot be eliminated, and particularly, in a dark light environment, in order to obtain a clear picture, the light sensitivity is generally increased, and the number of noise points is also increased. However, in the automatic driving simulation system, the image data acquired by the camera sensor is free of any noise, so that noise must be added to approach the photographing effect of the physical camera.
In the related art, image noise mainly includes gaussian noise and poisson noise. When noise is added to the simulation image, an algorithm for generating a random variable of a poisson distribution proposed by the gardner is generally adopted, but although the algorithm is simple, the calculation complexity is linear, and the calculation of a random number is time-consuming, so that the occupied GPU resources are very much, and the processing speed of the image is reduced. Further, Ahrens and Dieter present another improved algorithm in which for larger values of λ, either reject sampling, or a gaussian approximation of the poisson distribution is used. The amount of computation is still large, and even if the method is deployed on a GPU computing framework, the efficiency is not improved satisfactorily.
Based on this, the embodiment of the invention provides an image processing method in a driving simulation scene to solve the existing problems.
Referring to fig. 1, fig. 1 is a schematic structural diagram of an image processing system in a driving simulation scenario provided by an embodiment of the present invention, and to support an exemplary application, the image processing system in the driving simulation scenario includes a terminal (including a terminal 400-1 and a terminal 400-2) and a server 200, the terminal is connected to the server 200 through a network 300, and the network 300 may be a wide area network or a local area network, or a combination thereof, and implements data transmission using a wireless link.
The terminal (terminal 400-1 and/or terminal 400-2) is used for sending an image simulation request corresponding to the vehicle, wherein the image simulation request carries vehicle position parameters, environment parameters and exposure parameters;
in practical applications, the terminal may be various types of user terminals such as a smart phone, a tablet computer, a notebook computer, and the like, and may also be a wearable computing device, a Personal Digital Assistant (PDA), a desktop computer, a cellular phone, a media player, a navigation device, a game console, a television, or a combination of any two or more of these data processing devices or other data processing devices.
A server 200 for acquiring a simulation image output by an image rendering engine of a vehicle; determining the pixel value of each pixel point in the simulation image; generating random numbers corresponding to the pixel values based on the pixel values of the pixel points in the simulation image; searching and obtaining a Poisson noise value corresponding to each pixel point from the noise model based on the pixel value of each pixel point and the corresponding random number; generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point;
the server 200 is further configured to generate a simulation image obtaining request and send the simulation image obtaining request to the image rendering engine based on the received image simulation request of the corresponding vehicle.
In practical implementation, the server 200 may be a server configured separately to support various services, or may be a server cluster.
The following describes in detail a hardware structure of the image processing device in the driving simulation scenario provided by the embodiment of the present invention, referring to fig. 2, fig. 2 is a schematic structural diagram of the image processing device in the driving simulation scenario provided by the embodiment of the present invention, and it can be understood that fig. 2 only shows an exemplary structure of the image processing device in the driving simulation scenario, and not a whole structure thereof, and a part of the structure or a whole structure shown in fig. 2 may be implemented as required.
The image processing device in the driving simulation scene provided by the embodiment of the invention comprises: at least one processor 201, memory 202, user interface 203, and at least one network interface 204. The various components in the image processing apparatus 200 in the driving simulation scenario are coupled together by a bus system 205. It will be appreciated that the bus system 205 is used to enable communications among the components. The bus system 205 includes a power bus, a control bus, and a status signal bus in addition to a data bus. For clarity of illustration, however, the various buses are labeled as bus system 205 in fig. 2.
The user interface 203 may include, among other things, a display, a keyboard, a mouse, a trackball, a click wheel, a key, a button, a touch pad, or a touch screen.
It will be appreciated that the memory 202 can be either volatile memory or nonvolatile memory, and can include both volatile and nonvolatile memory. Among them, the nonvolatile Memory may be a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Programmable Read Only Memory (EPROM), a Flash Memory (Flash Memory), and the like. Volatile Memory can be Random Access Memory (RAM), which acts as external cache Memory. By way of example, and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM). The memory 202 described in connection with the embodiments of the invention is intended to comprise these and any other suitable types of memory.
The memory 202 in embodiments of the present invention is capable of storing data to support operation of the terminal (e.g., 400-1). Examples of such data include: any computer program, such as an operating system and application programs, for operating on a terminal (e.g., 400-1). The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, and the like, and is used for implementing various basic services and processing hardware-based tasks. The application program may include various application programs.
As an example of the implementation of the image processing device in the driving simulation scenario provided by the embodiment of the present invention by combining software and hardware, the image processing device in the driving simulation scenario provided by the embodiment of the present invention may be directly embodied as a combination of software modules executed by the processor 201, the software modules may be located in a storage medium, the storage medium is located in the memory 202, the processor 201 reads executable instructions included in the software modules in the memory 202, and the image processing method in the driving simulation scenario provided by the embodiment of the present invention is completed by combining necessary hardware (for example, including the processor 201 and other components connected to the bus 205).
By way of example, the Processor 201 may be an integrated circuit chip having Signal processing capabilities, such as a general purpose Processor, a Digital Signal Processor (DSP), or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or the like, wherein the general purpose Processor may be a microprocessor or any conventional Processor or the like.
As an example of the hardware implementation of the image processing Device in the driving simulation scenario provided by the embodiment of the present invention, the image processing Device in the driving simulation scenario provided by the embodiment of the present invention may be directly implemented by the processor 201 in the form of a hardware decoding processor, for example, by one or more Application Specific Integrated Circuits (ASICs), DSPs, Programmable Logic Devices (PLDs), Complex Programmable Logic Devices (CPLDs), Field Programmable Gate Arrays (FPGAs), or other electronic components to implement the method provided by the embodiment of the present invention.
The memory 202 in the embodiment of the present invention is used to store various types of data to support the operation of the image processing apparatus 200 in the driving simulation scene. Examples of such data include: any executable instructions for operating on the image processing apparatus 200 in the driving simulation scenario, such as executable instructions, may be included in the executable instructions, and the program implementing the image processing method in the driving simulation scenario of the embodiment of the present invention may be included in the executable instructions.
Based on the above description of the image processing system and apparatus in the driving simulation scene, the following description will discuss an image processing method in the driving simulation scene according to an embodiment of the present invention. Referring to fig. 3, fig. 3 is a schematic flow chart of an image processing method in a driving simulation scene provided in an embodiment of the present invention, in some embodiments, the image processing method in the driving simulation scene may be implemented by a server or a terminal, or implemented by the server and the terminal in a cooperative manner, taking the server as an example, the image processing method in the driving simulation scene provided in the embodiment of the present invention includes:
step 301: the server obtains a simulation image output by an image rendering engine of the vehicle.
In some embodiments, the server may obtain the simulated image output by the image rendering engine by:
the server sends a simulation image obtaining request to the image rendering engine, where the obtaining request carries image parameters used for generating a simulation image, and specifically, the image parameters may include a vehicle position parameter, an environment parameter, an exposure parameter, and the like.
After receiving the simulation image acquisition request, the image rendering engine analyzes the simulation image acquisition request to obtain image parameters carried in the simulation image acquisition request, performs image rendering according to the image parameters to obtain a simulation image and sends the simulation image to the server.
The server receives the simulation image, and the simulation image is not really shot because the simulation image is obtained from the image rendering engine, and does not carry any noise.
Step 302: and determining the pixel value of each pixel point in the simulation image.
In some embodiments, after the simulation image is obtained by the image rendering engine, the simulation image is processed to identify the pixel values of the respective pixel points in the simulation image.
In actual implementation, each pixel point can be traversed through multiple modes such as pointer scanning, iterator operation or dynamic address calculation based on an image processing library function provided by OpenCV, so as to obtain a pixel value of each pixel point in the simulation image.
Step 303: and generating random numbers corresponding to the pixel values based on the pixel values of the pixel points in the simulation image.
In some embodiments, it is necessary to generate, for each pixel point, a random number corresponding to the pixel value of the pixel point, so as to facilitate a subsequent poisson noise value corresponding to the pixel value in the noise model.
In practical implementation, the random number may be generated by directly using a random number generation function, such as a rand () function, a srand () function, and the like.
Because the random number exists to search for the poisson noise value at the corresponding position in the noise model, and the position of each poisson noise value in the noise model provided by the embodiment of the present invention is determined based on the random number set, the value range of the random number corresponding to each pixel value should be within the value range of the random number set identifying the position of each poisson noise value, or the same value range should be adopted. For example, if the value range of the random number set is [0, 2047], the random number corresponding to each pixel value may be 45, 1024, 1296, 2047, or the like.
Step 304: and searching and obtaining the Poisson noise value corresponding to each pixel point from the noise model based on the pixel value of each pixel point and the corresponding random number.
In some embodiments, before the image in the driving simulation scene is subjected to the noise adding process, a noise model needs to be constructed in advance, and the construction of the noise model provided by the embodiments of the present invention is described in detail below. Referring to fig. 4, fig. 4 is a schematic flowchart of noise model construction in a driving simulation scenario according to an embodiment of the present invention.
Step 304 a: and determining the gray value range and the noise level corresponding to the noise model.
Since camera image noise is usually an electronic noise representation, and is generally related to a camera sensor or its own circuit, when performing camera noise simulation, it is necessary to determine the noise level of the simulated noise according to some hardware-related parameters of the corresponding camera sensor, which may be 0.2 or 0.4, for example. Specifically, the accuracy of the camera sensor is high, and the corresponding noise level is lower, such as a camera sensor for industrial use.
In some embodiments, the noise model is constructed based on poisson noise. The poisson noise is generated by the fluctuation of the pixel gray value due to the light intensity, and therefore, when a noise model is constructed, a gray value range needs to be determined. Typically, the range of gray scale values is from 0 to 255, which can be employed in embodiments of the present invention.
Step 304 b: and determining the noise amplitude corresponding to each gray value belonging to the gray value range based on the gray value range and the noise level.
After determining the range of gray values and the noise level, the noise amplitude for determining the poisson noise value needs to be calculated. Since the noise amplitude of the poisson noise is related to the grey value, for each grey value belonging to the range of grey values a corresponding noise amplitude is determined.
In some embodiments, the noise amplitude corresponding to each gray value may be determined as follows: determining each gray value in the gray value range; and determining the noise amplitude corresponding to each gray value based on the inverse proportional relation between the preset gray value and the noise amplitude.
The range of gray values is typically [0, 255], with each gray value therein being determined, such as 0, 1, 2 … 255, based on the range of gray values. Since the darker the light source, the greater the pixel noise, the gray value is inversely proportional to the noise amplitude, and based on the inversely proportional relationship between the gray value and the noise amplitude, the noise amplitude corresponding to each gray value can be calculated by the following formula:
λ＝max(64-y,0)*N
where λ is the noise amplitude, y is the gray scale value, and N is the noise level.
Step 304 c: and generating a noise model based on the noise amplitude corresponding to each gray value.
In the embodiment of the present invention, the noise model is a two-dimensional data, and includes gray values and poisson noise values corresponding to the gray values.
In some embodiments, after determining the noise amplitude corresponding to each gray scale value in the above manner, the noise model may be generated according to the noise amplitude corresponding to each gray scale value by:
determining a noise value set corresponding to each gray value based on the noise amplitude corresponding to each gray value and the Poisson distribution relation established based on the noise amplitude; respectively acquiring random number sets corresponding to all gray values; respectively determining Poisson noise values corresponding to all random numbers in all random number sets based on the noise value sets and the random number sets corresponding to all gray values; and generating a noise model based on the Poisson noise value corresponding to each random number in each random number set.
Because the embodiment of the invention simulates the Poisson noise, after the noise amplitude corresponding to each gray value is determined, the noise value set corresponding to each gray value is calculated based on the noise amplitude and the constructed Poisson distribution relation.
In practical implementation, a value range of the corresponding noise value, that is, a noise amplitude range, may be determined according to the noise amplitude, for example, the noise amplitude is 7dB, and then the noise amplitude range is-7 dB to +7 dB. In performing the calculation of the set of noise values, each noise value may be generated by generating a random number within the range of noise amplitudes. Specifically, when generating the noise value set, the number of noise values may be determined according to the requirement of simulation accuracy, where the greater the number of noise values, the higher the accuracy of the corresponding noise values.
After the set of noise values is determined, the noise values are placed in the corresponding locations of the noise model so that it follows a poisson distribution. In some embodiments, a set of random numbers may be used to identify the location to which each noise value corresponds. The set of random numbers is for each gray value. And determining the Poisson noise value corresponding to each random number according to the random number set of each gray value, the corresponding noise value and the probability density function of Poisson distribution.
A noise model is generated based on the poisson noise values corresponding to the respective random numbers, and since the noise values in the noise model are subject to poisson distribution, the noise model may also be referred to as a poisson dictionary.
The following continues with the description of the image processing method in the driving simulation scenario in conjunction with the above-described noise model.
Since any color is composed of three primary colors of red, green and blue, the obtained pixel value of each pixel point is specifically the pixel values of each pixel point on three color channels of red R, green G and blue B. In view of the fact that the pixel value of each pixel point is obtained, and the noise model is constructed based on the range of the gray value of the pixel point, in some embodiments, the poisson noise value corresponding to each pixel point can be found from the noise model in the following manner:
determining gray values corresponding to the pixel points based on the pixel values of the pixel points on red R, green G and blue B color channels; and searching to obtain the Poisson noise value corresponding to each pixel point from the noise model based on the determined gray value and the random number corresponding to the pixel point.
In practical implementation, the Gray value RGB (Gray ) of each pixel point can be determined by the following method according to the pixel values RGB (R, G, B) of each pixel point on the three color channels of red R, green G, and blue B:
specifically, the conversion from pixel values to gray values is generally based on a weighted method, such as:
1) floating point arithmetic: gray ═ R0.3 + G0.59 + B0.11;
2) integer arithmetic: gray ═ (R30 + G59 + B11)/100;
3) and (3) a shifting algorithm: gray ═ (R28 + G151 + B77) > > 8;
4) average value method: gray ═ R + G + B)/3.
After the Gray value is obtained by any one of the above methods, R, G, B in the original RGB (R, G, B) are replaced by the Gray value in a unified manner, so as to obtain the Gray value RGB (Gray ) of each pixel point.
After the gray value of each pixel point is determined, aiming at each pixel point, according to the gray value of the pixel point, a Poisson noise value sequence corresponding to the gray value is searched in a noise model; and then according to the random number corresponding to the pixel point, finding out the Poisson noise value at the corresponding position in the determined Poisson noise value sequence.
Since the noise model can be stored as image data, it can be conveniently used as a texture resource of the GPU. The GPU provides a buffer for texture operation specially, and has high access performance, so that when the Poisson noise value is searched, the Poisson noise value corresponding to each pixel point can be searched from the noise model through the graphics processor, and the image processing efficiency is improved.
Step 305: and generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point.
And processing the pixel value of each pixel point in the acquired simulation image based on the Poisson noise value corresponding to each pixel point searched in the noise model, and generating a noise image for driving simulation based on the processed pixel value of each pixel point.
In some embodiments, each pixel point may be processed to generate a noisy image for driving simulation by: respectively determining the sum of the pixel value of each pixel point and the corresponding Poisson noise value to obtain a first target pixel value corresponding to each pixel point; and generating a noise image for driving simulation based on the first target pixel value of each pixel point.
In actual implementation, since the noise model is stored as image data, when adding noise to the simulation image, the acquired poisson noise value corresponding to each pixel point may be added to the corresponding pixel value to obtain a first target pixel value corresponding to each pixel point, and a noise image for driving simulation may be generated based on the first target pixel value.
In addition, based on the above description, it can be seen that the noise model is constructed based on poisson noise, and therefore, the simulation of poisson noise is realized by adopting the above steps, the reason for forming the noise point of the image shot by the real camera is complicated, and the types of generated noise may be various. In some embodiments, in addition to poisson noise, gaussian noise may be added to the simulated image.
In some embodiments, gaussian noise may be generated as follows: and determining the Gaussian noise value corresponding to each pixel point based on the random number corresponding to each pixel point.
The gaussian noise is a type of noise whose probability density function follows gaussian distribution, and in specific implementation, there are various methods for simulating generation of gaussian noise, such as an inverse function of a standard normal cumulative distribution function, Box-Muller transform, ziggurat algorithm, and the like. And obtaining the Gaussian noise value of which the noise value obeys Gaussian distribution by the mode according to the random number corresponding to each pixel point.
Based on this, in some embodiments, a noise image for driving simulation may also be generated as follows: respectively determining the summation of the pixel value of each pixel point and the corresponding noise value and Gaussian noise value to obtain a second target pixel value corresponding to each pixel point; and generating a noise image for driving simulation based on the second target pixel value of each pixel point.
The method for adding the Gaussian noise to the simulated image and the method for adding the Poisson noise can adopt the same method, specifically, the pixel value of each pixel point of the simulated image can be directly added with the Gaussian noise value, and the first target pixel value added with the Poisson noise value can also be added with the Gaussian noise value, so that the second target pixel value of each pixel point added with the Gaussian noise can be obtained; and generating a noise image for driving simulation based on the second target pixel value of each pixel point.
Next, a description is continued on an image processing method in a driving simulation scene provided by an embodiment of the present invention, and referring to fig. 5, fig. 5 is a schematic flow chart of adding poisson noise to a simulation image provided by an embodiment of the present invention.
Step 501: and sending a simulation image acquisition request to an image rendering engine to acquire a simulation image.
Here, the simulation image acquisition request may carry image parameters for simulation image generation, such as vehicle position parameters, environmental parameters, exposure parameters, and the like.
Step 502: and identifying each pixel point in the simulation image, and determining the pixel value of each pixel point.
Step 503: based on the pixel values of the pixels, random numbers corresponding to the pixel values are generated.
Here, the random number may be generated using a random number generation function for inquiring a poisson noise value corresponding to each pixel value in the noise model, and a range of the random number corresponds to a range of X-dimension in the noise model.
Step 504: and searching the Poisson noise value corresponding to each pixel point in the noise model based on the pixel value and the random number of each pixel point.
Here, the poisson noise values in the noise model follow a poisson distribution; the noise model is pre-constructed, which is built on the basis of gray values, and each gray value corresponds to a poisson noise value that follows a poisson distribution, since the generation of poisson noise is related to the light intensity, which directly affects the pixel gray values of the image.
In addition, since the obtained pixel value of each pixel point is obtained, when the noise model is queried, the pixel value needs to be converted into a gray value by using the weighting method, and then the corresponding poisson noise value is searched according to the gray value of each pixel point.
Step 505: and adding the pixel value of each pixel point and the corresponding Poisson noise value to obtain a first target pixel value of each pixel point.
Step 506: a noise image for driving simulation is generated based on the first target pixel value.
The noise image is obtained by adding Poisson noise to the acquired simulation image, and the simulation of the Poisson noise in the driving simulation scene is realized.
Next, a description is continued on an image processing method in a driving simulation scene provided by an embodiment of the present invention, referring to fig. 6, where fig. 6 is a schematic flow chart illustrating a process of adding gaussian noise to a simulation image provided by an embodiment of the present invention.
Step 601: the simulation image is acquired by sending a simulation image acquisition request to the image rendering engine.
Step 602: and identifying each pixel point in the simulation image, and determining the pixel value of each pixel point.
Step 603: based on the pixel values of the pixels, random numbers corresponding to the pixel values are generated.
Step 604: and searching the Poisson noise value corresponding to each pixel point in the noise model based on the pixel value and the random number of each pixel point.
Step 605: and determining the Gaussian noise value corresponding to each pixel point based on the random number corresponding to each pixel point.
Here, a corresponding noise value is generated by using a Bo x-Muller transform or the like based on the random number corresponding to each pixel point and the probability density function of the gaussian distribution, and the noise value is a gaussian noise value subject to the gaussian distribution.
Step 606: and adding the pixel value of each pixel point with the corresponding Poisson noise value and Gaussian noise value to obtain a second target pixel value of each pixel point.
Here, when calculating the second target pixel value of each pixel, the pixel value of each pixel is calculated,
Step 607: a noise image for driving simulation is generated based on the second target pixel value.
Here, the noise image is obtained by adding poisson noise and gaussian noise to the acquired simulation image at the same time. It should be noted that, when the simulation image is processed here, only poisson noise may be added, such as the above step 501-506, only gaussian noise may be added, or both poisson noise and gaussian noise may be added, which may be determined according to the requirement for the driving simulation scene.
Continuing to describe the image processing method in the driving simulation scene provided by the embodiment of the present invention, referring to fig. 7, fig. 7 is a schematic flow chart of the image processing method in the driving simulation scene provided by the embodiment of the present invention, and the image processing method in the driving simulation scene provided by the embodiment of the present invention includes:
step 701: and acquiring a simulation image output by the image rendering engine.
Step 702: and determining the pixel value of each pixel point in the simulation image.
Step 703: and inputting the pixel value of each pixel point and the corresponding random number to a Poisson dictionary, and inquiring the Poisson noise value corresponding to each pixel point.
Here, the poisson dictionary is the noise model described above. The Poisson dictionary is two-dimensional data, can be stored as a single-channel gray-scale image and is deployed on a GPU. Including X and Y dimensions, where the range of the X dimension is free, representing a range of random numbers corresponding to each pixel value, the random numbers identifying the location of each poisson noise value in the poisson dictionary, such as [0, 2047 ]; the Y dimension is a range of gray scale values of pixel values, such as [0, 255 ].
In the process of generating the poisson dictionary, the following algorithm program can be adopted to implement:
let D [256] [2048] ← 0, p ← 0, s ← p; generating a poisson dictionary of 256 × 2048 size, with an initial value of poisson noise value of 0.
For y←0to 255do：
Letλ←max(64-y，0)*N，p←e
-λS ← p; // determining the noise amplitude based on the gray value and the noise level.
For k←1toλdo：
p ← p x λ/k; // Poisson noise value corresponding to the probability of the Poisson distribution.
s ← s + p; // Poisson noise value corresponding to the cumulative probability.
D [ y ] [ int(s) × 2048] ← k-1; the// poisson noise value corresponds to the position in the poisson dictionary.
For j←0to 2047do：
If j>0and D[y][j]＝0：
D [ y ] [ j ] ← D [ y ] [ j-1 ]; fill in Poisson noise values at the location in Poisson dictionary where the noise value is 0.
D [ y ] [ j ] </D [ y ] [ j ] -lambda/2.// modifying the Poisson noise value to fit the noise amplitude range.
Wherein, λ is the noise amplitude, y is the gray value, N is the Poisson noise level, D [256] [2048] is the size of the Poisson dictionary, and p is the Poisson distribution probability corresponding to each Poisson noise value.
Specifically, based on the algorithm program, a two-dimensional data matrix of the poisson dictionary 256 × 2048 is generated, the X dimension is a random number set of [0, 2047], and the Y dimension is each gray value set within the range of [0, 255 ].
When the Poisson dictionary is inquired, the pixel value corresponding to each pixel point can be converted into a gray value, and the Poisson noise value corresponding to the pixel point is inquired based on the gray value and the corresponding random number.
Step 704: and generating a Gaussian noise value corresponding to each pixel point based on the random number corresponding to each pixel point.
Step 705: and summing the pixel value of each pixel point, the corresponding Poisson noise value and the Gaussian noise value to obtain a target pixel value corresponding to each pixel point, and generating a noise image for driving simulation based on the target pixel value.
Here, referring to fig. 8, fig. 8 is a schematic flowchart of implementing noise simulation based on a poisson dictionary according to an embodiment of the present invention. Specifically, based on the pixel value of each pixel point and the corresponding random number, the corresponding Poisson noise value is searched in a Poisson dictionary; then generating a corresponding Gaussian noise value based on the random number of each pixel point; and adding the pixel value of each pixel point, the Gaussian noise value and the Poisson noise value to obtain a final target pixel value for generating a noise image of the driving simulation scene.
The poisson dictionary is preloaded on the GPU, and the server performs noise adding processing on the simulation image through the GPU.
Generating a random number corresponding to each pixel value by acquiring the pixel value of each pixel point in the simulation image, searching a Poisson noise value corresponding to each pixel point in a preset noise model based on the pixel value and the corresponding random number, and adding the Poisson noise value and the pixel value of each corresponding pixel point to obtain a noise image for driving simulation; because the noise model is pre-constructed aiming at the pixel value of each pixel point, the Poisson noise value corresponding to each pixel value can be directly searched in the noise model when the image processing is carried out so as to realize the noise adding processing of the simulation image, thereby reducing the calculation amount and improving the generation efficiency of the noise image.
Continuing with the description of the image processing device in the driving simulation scenario provided by the embodiments of the present invention, in some embodiments, the image processing device in the driving simulation scenario may be implemented by using a software module. Referring to fig. 9, fig. 9 is a schematic structural diagram of an image processing apparatus in a driving simulation scene according to an embodiment of the present invention, where the image processing apparatus 900 in the driving simulation scene according to the embodiment of the present invention includes:
an obtaining unit 901 configured to obtain a simulation image output by an image rendering engine of a vehicle;
a determining unit 902, configured to determine a pixel value of each pixel in the simulated image;
a first generating unit 903, configured to generate a random number corresponding to each pixel value based on the pixel value of each pixel point in the simulation image;
a searching unit 904, configured to search for a poisson noise value corresponding to each pixel point from a noise model based on the pixel value of each pixel point and a corresponding random number, where the noise value in the noise model obeys poisson distribution;
a second generating unit 905, configured to generate a noise image for driving simulation based on the pixel value of each pixel in the simulation image and the poisson noise value corresponding to each pixel.
In some embodiments, further comprising:
a model building unit 906, configured to determine a gray value range and a noise level corresponding to the noise model;
determining the noise amplitude corresponding to each gray value belonging to the gray value range based on the gray value range and the noise level;
and generating the noise model based on the noise amplitude corresponding to each gray value.
In some embodiments, the model construction unit 906 is configured to determine respective gray values in the gray value range;
and determining the noise amplitude corresponding to each gray value based on the inverse proportional relation between the preset gray value and the noise amplitude.
In some embodiments, the model constructing unit 906 is configured to determine a noise value set corresponding to each gray value based on a noise amplitude corresponding to each gray value and a poisson distribution relationship constructed based on the noise amplitude;
respectively acquiring a random number set corresponding to each gray value;
respectively determining Poisson noise values corresponding to all random numbers in all random number sets based on the noise value sets and the random number sets corresponding to all gray values;
and generating the noise model based on the Poisson noise value corresponding to each random number in each random number set.
In some embodiments, the searching unit 904 is configured to determine a gray value corresponding to the pixel point based on pixel values of the pixel point on red R, green G, and blue B color channels;
and searching to obtain the Poisson noise value corresponding to each pixel point from the noise model based on the determined gray value and the random number corresponding to the pixel point.
In some embodiments, the second generating unit 905 is configured to determine a sum of a pixel value of each pixel and a corresponding poisson noise value, respectively, to obtain a first target pixel value corresponding to each pixel;
and generating a noise image for driving simulation based on the first target pixel value of each pixel point.
In some embodiments, the searching unit 904 is configured to determine a gaussian noise value corresponding to each pixel point based on the random number corresponding to each pixel point.
In some embodiments, the second generating unit 905 is configured to determine the sum of the pixel value of each pixel and the corresponding noise value and gaussian noise value, respectively, to obtain a second target pixel value corresponding to each pixel;
and generating a noise image for driving simulation based on the second target pixel value of each pixel point.
In some embodiments, the searching unit 904 is configured to search, by the graphics processor, a poisson noise value corresponding to each pixel point from the noise model based on the pixel value of each pixel point and the corresponding random number.
Here, it should be noted that: the above description related to the image processing device in the driving simulation scene is similar to the above description of the method, and the description of the beneficial effects of the method is not repeated, and for the technical details not disclosed in the image processing device in the driving simulation scene according to the embodiment of the present invention, please refer to the description of the embodiment of the method of the present invention.
The embodiment of the invention also provides an image processing device in the driving simulation scene, which comprises the following components:
a memory for storing executable instructions;
and the processor is used for realizing the image processing method in the driving simulation scene provided by the embodiment of the invention when the executable instructions stored in the memory are executed.
The embodiment of the invention also provides a storage medium, wherein the storage medium stores executable instructions for causing a processor to execute the executable instructions so as to realize the image processing method in the driving simulation scene provided by the embodiment of the invention.
In some embodiments, the storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EE PROM, flash, magnetic surface memory, optical disk, or CD-ROM; or may be various devices including one or any combination of the above memories.
In some embodiments, executable instructions may be written in any form of programming language (including compiled or interpreted languages), in the form of programs, software modules, scripts or code, and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.
By way of example, executable instructions may correspond, but do not necessarily have to correspond, to files in a file system, may be stored in a portion of a file that holds other programs or data, e.g., in one or more scripts in a HyperText markup Language (H TML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code).
By way of example, executable instructions may be deployed to be executed on one computing device or on multiple computing devices at one site or distributed across multiple sites and interconnected by a communication network.
The above description is only an example of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present invention are included in the protection scope of the present invention.
Claims (10)
1. A method of image processing in a driving simulation scenario, the method comprising:
acquiring a simulation image output by an image rendering engine of a vehicle;
determining the pixel value of each pixel point in the simulation image;
generating a random number corresponding to each pixel value based on the pixel value of each pixel point in the simulation image;
based on the pixel value of each pixel point and the corresponding random number, searching a noise model to obtain a Poisson noise value corresponding to each pixel point, wherein the noise value in the noise model obeys Poisson distribution;
and generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point.
2. The method of claim 1, wherein the method further comprises:
determining a gray value range and a noise level corresponding to the noise model;
determining the noise amplitude corresponding to each gray value belonging to the gray value range based on the gray value range and the noise level;
and generating the noise model based on the noise amplitude corresponding to each gray value.
3. The method of claim 2, wherein determining the noise amplitude corresponding to each gray scale value belonging to the gray scale value range based on the gray scale value range and the noise level comprises:
determining each gray value in the gray value range;
and determining the noise amplitude corresponding to each gray value based on the inverse proportional relation between the preset gray value and the noise amplitude.
4. The method of claim 2, wherein generating the noise model based on the noise magnitude corresponding to each of the gray scale values comprises:
determining a noise value set corresponding to each gray value based on the noise amplitude corresponding to each gray value and the Poisson distribution relation constructed based on the noise amplitude;
respectively acquiring a random number set corresponding to each gray value;
respectively determining Poisson noise values corresponding to all random numbers in all random number sets based on the noise value sets and the random number sets corresponding to all gray values;
and generating the noise model based on the Poisson noise value corresponding to each random number in each random number set.
5. The method of claim 1, wherein the pixel values of the pixel points comprise: pixel values of the pixel points on red R, green G and blue B color channels respectively;
the searching for the poisson noise value corresponding to each pixel point from the noise model based on the pixel value of each pixel point and the corresponding random number comprises the following steps:
determining gray values corresponding to the pixel points based on the pixel values of the pixel points on red R, green G and blue B color channels;
and searching to obtain the Poisson noise value corresponding to each pixel point from the noise model based on the determined gray value and the random number corresponding to the pixel point.
6. The method of claim 1, wherein generating a noise image for driving simulation based on pixel values of pixels in the simulation image and poisson noise values corresponding to pixels comprises:
respectively determining the sum of the pixel value of each pixel point and the corresponding Poisson noise value to obtain a first target pixel value corresponding to each pixel point;
and generating a noise image for driving simulation based on the first target pixel value of each pixel point.
7. The method of claim 1, wherein the method further comprises:
determining a Gaussian noise value corresponding to each pixel point based on the random number corresponding to each pixel point;
the generating of the noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the noise value corresponding to each pixel point includes:
respectively determining the summation of the pixel value of each pixel point and the corresponding noise value and Gaussian noise value to obtain a second target pixel value corresponding to each pixel point;
and generating a noise image for driving simulation based on the second target pixel value of each pixel point.
8. The method of any one of claims 1 to 7, wherein the finding a poisson noise value corresponding to each pixel point from a noise model based on the pixel value of each pixel point and a corresponding random number comprises:
and searching the Poisson noise value corresponding to each pixel point from the noise model through the graphic processor based on the pixel value of each pixel point and the corresponding random number.
9. An image processing apparatus in a driving simulation scenario, the apparatus comprising:
the system comprises an acquisition unit, a display unit and a control unit, wherein the acquisition unit is used for acquiring a simulation image output by an image rendering engine of a vehicle;
the determining unit is used for determining the pixel value of each pixel point in the simulation image;
the first generation unit is used for generating random numbers corresponding to pixel values of all pixels in the simulation image based on the pixel values of all the pixels in the simulation image;
the searching unit is used for searching and obtaining a Poisson noise value corresponding to each pixel point from a noise model based on the pixel value of each pixel point and the corresponding random number, and the noise value in the noise model obeys Poisson distribution;
and the second generation unit is used for generating a noise image for driving simulation based on the pixel value of each pixel point in the simulation image and the Poisson noise value corresponding to each pixel point.
10. A storage medium characterized by storing executable instructions for implementing the image processing method in the driving simulation scenario of any one of claims 1 to 8 when executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910850744.5A CN110782391B (en) | 2019-09-10 | 2019-09-10 | Image processing method and device in driving simulation scene and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910850744.5A CN110782391B (en) | 2019-09-10 | 2019-09-10 | Image processing method and device in driving simulation scene and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110782391A true CN110782391A (en) | 2020-02-11 |
| CN110782391B CN110782391B (en) | 2022-06-03 |
Family
ID=69383417
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910850744.5A Active CN110782391B (en) | 2019-09-10 | 2019-09-10 | Image processing method and device in driving simulation scene and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110782391B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111597707A (en) * | 2020-05-14 | 2020-08-28 | 北京百度网讯科技有限公司 | Processing method, device and equipment of simulation scene and storage medium |
| CN112672052A (en) * | 2020-12-24 | 2021-04-16 | 华中光电技术研究所（中国船舶重工集团公司第七一七研究所） | Image data enhancement method and system, electronic equipment and storage medium |
| CN114049425A (en) * | 2021-10-29 | 2022-02-15 | 腾讯科技（深圳）有限公司 | Method, apparatus, device, medium and program product for simulating illumination in an image |
| CN114972396A (en) * | 2022-06-17 | 2022-08-30 | 常州企智略科技有限公司 | Image edge positioning and denoising technology based on gray detection |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1969297A (en) * | 2001-06-15 | 2007-05-23 | 索尼公司 | Image processing apparatus and method and image pickup apparatus |
| JP2013179542A (en) * | 2012-02-29 | 2013-09-09 | Nikon Corp | Image processing device and program |
| CN105718235A (en) * | 2016-03-22 | 2016-06-29 | 华为技术有限公司 | Image display processing method, image display processing device and image display system |
| CN109255772A (en) * | 2018-08-27 | 2019-01-22 | 平安科技（深圳）有限公司 | License plate image generation method, device, equipment and medium based on Style Transfer |
| CN109308380A (en) * | 2018-08-24 | 2019-02-05 | 云南大学 | Embroidery art style analogy method based on feeling of unreality |
| CN109643125A (en) * | 2016-06-28 | 2019-04-16 | 柯尼亚塔有限公司 | For training the 3D virtual world true to nature of automated driving system to create and simulation |
| CN109902806A (en) * | 2019-02-26 | 2019-06-18 | 清华大学 | Determination method of target bounding box of noisy image based on convolutional neural network |
| CN110070515A (en) * | 2019-01-31 | 2019-07-30 | 北京字节跳动网络技术有限公司 | Image composition method, device and computer readable storage medium |
-
2019
- 2019-09-10 CN CN201910850744.5A patent/CN110782391B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1969297A (en) * | 2001-06-15 | 2007-05-23 | 索尼公司 | Image processing apparatus and method and image pickup apparatus |
| JP2013179542A (en) * | 2012-02-29 | 2013-09-09 | Nikon Corp | Image processing device and program |
| CN105718235A (en) * | 2016-03-22 | 2016-06-29 | 华为技术有限公司 | Image display processing method, image display processing device and image display system |
| CN109643125A (en) * | 2016-06-28 | 2019-04-16 | 柯尼亚塔有限公司 | For training the 3D virtual world true to nature of automated driving system to create and simulation |
| CN109308380A (en) * | 2018-08-24 | 2019-02-05 | 云南大学 | Embroidery art style analogy method based on feeling of unreality |
| CN109255772A (en) * | 2018-08-27 | 2019-01-22 | 平安科技（深圳）有限公司 | License plate image generation method, device, equipment and medium based on Style Transfer |
| CN110070515A (en) * | 2019-01-31 | 2019-07-30 | 北京字节跳动网络技术有限公司 | Image composition method, device and computer readable storage medium |
| CN109902806A (en) * | 2019-02-26 | 2019-06-18 | 清华大学 | Determination method of target bounding box of noisy image based on convolutional neural network |
Non-Patent Citations (3)
| Title |
|---|
| R. TANAKA ET AL: "Noise simulation system for determining imaging conditions in digital radiography", 《MEDICAL IMAGING 2012: PHYSICS OF MEDICAL IMAGING》 * |
| 张智恒: "虚拟环境下相机成像的噪声模拟", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王彦等: "三维动态红外视景仿真中噪声的模拟", 《激光与红外》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111597707A (en) * | 2020-05-14 | 2020-08-28 | 北京百度网讯科技有限公司 | Processing method, device and equipment of simulation scene and storage medium |
| CN111597707B (en) * | 2020-05-14 | 2023-09-29 | 北京百度网讯科技有限公司 | Method, device, equipment and storage medium for processing simulation scene |
| CN112672052A (en) * | 2020-12-24 | 2021-04-16 | 华中光电技术研究所（中国船舶重工集团公司第七一七研究所） | Image data enhancement method and system, electronic equipment and storage medium |
| CN114049425A (en) * | 2021-10-29 | 2022-02-15 | 腾讯科技（深圳）有限公司 | Method, apparatus, device, medium and program product for simulating illumination in an image |
| CN114049425B (en) * | 2021-10-29 | 2023-06-09 | 腾讯科技（深圳）有限公司 | Illumination simulation method, device, equipment and storage medium in image |
| CN114972396A (en) * | 2022-06-17 | 2022-08-30 | 常州企智略科技有限公司 | Image edge positioning and denoising technology based on gray detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110782391B (en) | 2022-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110782391B (en) | Image processing method and device in driving simulation scene and storage medium | |
| CN112233216B (en) | Game image processing method and device and electronic equipment | |
| JPH10510074A (en) | Image composition | |
| CN114119853B (en) | Image rendering method, apparatus, device and medium | |
| CN111583381B (en) | Game resource map rendering method and device and electronic equipment | |
| CN112712582A (en) | Dynamic global illumination method, electronic device and computer-readable storage medium | |
| CN110917617B (en) | Method, device, equipment and storage medium for generating water ripple image | |
| US20230033319A1 (en) | Method, apparatus and device for processing shadow texture, computer-readable storage medium, and program product | |
| CN114792359A (en) | Rendering network training and virtual object rendering method, device, equipment and medium | |
| CN111915712A (en) | Illumination rendering method and device, computer readable medium and electronic equipment | |
| CN114742931A (en) | Method and device for rendering image, electronic equipment and storage medium | |
| CN117372607A (en) | Three-dimensional model generation method and device and electronic equipment | |
| CN116415652A (en) | Data generation method and device, readable storage medium and terminal equipment | |
| CN115512025A (en) | Method and device for detecting model rendering performance, electronic device and storage medium | |
| CN113516697A (en) | Method, apparatus, electronic device, and computer-readable storage medium for image registration | |
| CN113724364B (en) | Setting method and device for realizing shielding and rendering-free body by utilizing polygons | |
| CN117237514A (en) | Image processing method and image processing apparatus | |
| CN113786616B (en) | Indirect illumination implementation method and device, storage medium and computing equipment | |
| CN114782666A (en) | Recognition model training method and device and license plate recognition method and device | |
| US20060250414A1 (en) | System and method of anti-aliasing computer images | |
| CN113744379A (en) | Image generation method and device and electronic equipment | |
| KR102687528B1 (en) | Method and system for deleting text in image | |
| CN114494384B (en) | Building shadow analysis method, device, equipment and storage medium | |
| CN117611727B (en) | Rendering processing method, device, equipment and medium | |
| CN114288650B (en) | Virtual scene image processing method, display method, device, and electronic device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |