CN114359514A

CN114359514A - Binocular stereo image parallax image acquisition method, device, equipment and medium

Info

Publication number: CN114359514A
Application number: CN202111664641.3A
Authority: CN
Inventors: 陈伟海; 赵小铭; 张建斌; 孙先涛; 陈文杰
Original assignee: Beihang University; Anhui University
Current assignee: Beihang University; Anhui University
Priority date: 2021-12-30
Filing date: 2021-12-30
Publication date: 2022-04-15

Abstract

The embodiment of the specification discloses a binocular stereo image parallax image acquisition method, which comprises the following steps: acquiring a binocular stereo image to be processed and laser radar point cloud; performing feature extraction on the binocular stereo images to determine a feature correlation diagram corresponding to the left image and the right image in the binocular stereo images; generating a feature correlation pyramid according to the feature correlation graph corresponding to the left image and the right image; projecting the laser radar point cloud to an image on the appointed side in the binocular stereo image to generate a related laser radar parallax image; extracting an image on the appointed side in the binocular stereo image to determine an affinity propagation diagram and context characteristics; generating a parallax map according to the affinity propagation map and the laser radar parallax image; searching corresponding relevant features in the feature relevant pyramid according to the disparity map; and inputting the related features and the context features into a GRU updating module, and obtaining the binocular stereo parallax image in an iterative manner.

Description

Binocular stereo image parallax image acquisition method, device, equipment and medium

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a medium for acquiring a binocular stereoscopic parallax image.

Background

Binocular Stereo Vision (Binocular Stereo Vision) is an important form of machine Vision, and is a method for acquiring three-dimensional geometric information of an object by acquiring two images of the object to be measured from different positions by using imaging equipment based on a parallax principle and calculating position deviation between corresponding points of the images. The binocular stereo vision integrates images obtained by two eyes and observes the difference between the images, so that obvious depth sense can be obtained, the corresponding relation between the characteristics is established, and mapping points of the same space physical point in different images are corresponded, and the difference is called parallax (Disparity) image.

In the prior art, when disparity matching of binocular stereo images is performed, interference of an external environment is caused, and particularly, when the disparity matching of the binocular stereo images is performed in an outdoor environment, factors such as sunlight interfere with the disparity matching of the binocular stereo images, so that finally generated disparity images are possibly inaccurate, and a final result is influenced.

Disclosure of Invention

One or more embodiments of the present specification provide a binocular stereoscopic image parallax image acquisition method, apparatus, device, and medium, which are used to solve the following technical problems:

when the disparity of the binocular stereo image is matched, the disparity of the binocular stereo image can be interfered by the external environment, particularly in the outdoor environment, the disparity matching of the binocular stereo image can be interfered by factors such as sunlight, and further the finally generated disparity image can be inaccurate, so that the final result is influenced.

One or more embodiments of the present disclosure adopt the following technical solutions:

one or more embodiments of the present specification provide a binocular stereoscopic image parallax image acquiring method, including:

acquiring a binocular stereo image to be processed and laser radar point cloud, wherein the binocular stereo image comprises a left image and a right image which are shot by a left camera and a right camera;

performing feature extraction on the binocular stereo image to determine a feature correlation diagram corresponding to a left image and a right image in the binocular stereo image;

generating a feature correlation pyramid according to the feature correlation graph corresponding to the left image and the right image;

projecting the laser radar point cloud to an image on the appointed side in the binocular stereo image to generate a related laser radar parallax image;

extracting an image on the appointed side in the binocular stereo image to determine an affinity propagation diagram and context characteristics;

generating a disparity map according to the affinity propagation map and the laser radar disparity image, and searching corresponding relevant features in the feature-related pyramid according to the disparity map;

and inputting the related features and the context features into a GRU updating module to obtain a binocular stereo parallax image in an iterative manner.

Further, the performing feature extraction on the binocular stereo image to determine a feature correlation diagram corresponding to the left image and the right image in the binocular stereo image specifically includes:

inputting the binocular stereo image into a feature extraction network;

extracting a feature correlation diagram corresponding to a left image and a right image in the binocular stereo image through the feature extraction network; the feature extraction network is composed of a residual error module and a down-sampling layer.

Further, the extracting the image on the specified side in the binocular stereo image to determine an affinity propagation diagram and context features specifically includes:

inputting an image at a specified side in the binocular stereo image into a context extraction network;

and extracting the affinity propagation graph and the context characteristics of the image on the appointed side in the binocular stereo image through the context extraction network.

Furthermore, a plurality of GRU updating modules are arranged;

the inputting the relevant features and the contextual features into a GRU updating module to obtain a binocular stereo parallax image in an iterative manner specifically includes:

and inputting the relevant features and the context features into a GRU updating module, and obtaining the binocular stereo parallax image through iterative calculation of a plurality of GRU updating modules.

Further, the inputting the relevant features and the contextual features into a GRU update module, and obtaining the binocular stereoscopic parallax image through iterative computation by the plurality of GRU update modules specifically includes:

and taking the feature of the image on the appointed side as a first hidden feature, inputting the first hidden feature, the related feature and the context feature into a GRU updating module, outputting the updated hidden feature and parallax variable quantity, adding the parallax variable quantity and the parallax map, performing parallax propagation on the added parallax by using an affinity propagation map and a laser radar parallax map to obtain an iterated parallax map, and performing iterative processing according to the rest GRU updating modules to obtain a binocular stereo parallax image.

Further, the loss function of the binocular stereo image parallax matching model is one or more of a sparse parallax loss function, a left-right consistent loss function and a smooth loss function.

Further, the resolution of the feature correlation map is 1/4 or 1/8 of the binocular stereo image.

One or more embodiments of the present specification provide a binocular stereoscopic image parallax image acquiring apparatus including:

the system comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring a binocular stereo image to be processed and laser radar point cloud, and the binocular stereo image comprises a left side image and a right side image which are shot by a left camera and a right camera;

the extraction unit is used for extracting the features of the binocular stereo images and determining a feature correlation diagram corresponding to a left image and a right image in the binocular stereo images;

the first generation unit is used for generating a feature correlation pyramid according to the feature correlation graph corresponding to the left image and the right image;

the second generation unit is used for projecting the laser radar point cloud to an image on the appointed side in the binocular stereo image to generate a related laser radar parallax image;

the determining unit is used for extracting an image on the appointed side in the binocular stereo image and determining an affinity propagation diagram and context characteristics;

the third generating unit is used for generating a disparity map according to the affinity propagation map and the laser radar disparity image, and searching corresponding relevant features in the feature-related pyramid according to the disparity map;

and the acquisition unit is used for inputting the related features and the contextual features into the GRU updating module to obtain the binocular stereo parallax image in an iterative manner.

at least one processor; and the number of the first and second groups,

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to:

One or more embodiments of the present specification provide a non-transitory computer storage medium storing computer-executable instructions configured to:

The embodiment of the specification adopts at least one technical scheme which can achieve the following beneficial effects:

the method comprises the steps of simultaneously acquiring a binocular stereo image to be processed and laser radar point cloud, and performing feature extraction on the binocular stereo image to obtain a feature correlation diagram of a left side image and a right side image so as to generate a correlation pyramid through the feature correlation diagram in a subsequent step; meanwhile, the laser radar point cloud is projected to an image on the appointed side in the binocular stereo image to generate a related laser radar parallax image, the image on the appointed side in the binocular stereo image is extracted subsequently to determine an affinity propagation diagram and context characteristics, then the parallax image is generated according to the affinity propagation diagram and the laser radar parallax image, corresponding related characteristics are searched in a characteristic related pyramid according to the parallax image, and finally the related characteristics and the context characteristics are input into a GRU updating module to obtain the binocular stereo parallax image in an iterative mode.

Drawings

In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative effort. In the drawings:

fig. 1 is a schematic flowchart of a binocular stereo image parallax image acquiring method according to one or more embodiments of the present disclosure;

fig. 2 is a schematic structural diagram of a binocular stereo image disparity matching model provided in one or more embodiments of the present disclosure;

fig. 3 is a schematic structural diagram of a binocular stereoscopic image parallax image acquiring apparatus according to one or more embodiments of the present disclosure;

fig. 4 is a schematic structural diagram of a binocular stereoscopic image parallax image acquiring apparatus according to one or more embodiments of the present disclosure.

Detailed Description

The embodiment of the specification provides a binocular stereo image parallax image acquisition method, device, equipment and medium.

When the disparity of the binocular stereo image is matched, the disparity of the binocular stereo image can be interfered by the external environment, particularly, when the disparity of the binocular stereo image is matched in an outdoor environment, external factors such as sunlight can interfere the disparity matching of the binocular stereo image, and therefore the finally generated disparity image can be inaccurate, and the final result is influenced.

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present specification, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present specification without any creative effort shall fall within the protection scope of the present specification.

Fig. 1 is a schematic flowchart of a method for acquiring a binocular stereo image parallax image according to one or more embodiments of the present disclosure, where the method may be applied to a binocular stereo image parallax matching system, and the system may be combined with the application of binocular stereo image parallax, so that a binocular stereo image parallax matching model constructed by the system may better generate a parallax image of a binocular stereo image. Certain input parameters or intermediate results in the flow allow for manual intervention adjustments to help improve accuracy.

The method of the embodiment of the specification comprises the following steps:

s102, acquiring a binocular stereo image to be processed and a laser radar point cloud, wherein the binocular stereo image comprises a left image and a right image which are shot by a left camera and a right camera.

In the embodiment of the specification, the laser radar image and the binocular stereo image are shot at the same time and at the same angle in the same scene.

Because the existing binocular stereo image is limited by the environment when acquiring data, especially in the outdoor environment, most of the existing binocular stereo image is used for synthesizing a data set indoors to train a supervision network, and the generalization capability of the network is greatly limited. However, the lidar point cloud can reduce the influence of the environment as much as possible. Therefore, the self-supervision training by using the binocular stereo image and the laser radar point cloud can well solve the problems.

And S104, performing feature extraction on the binocular stereo image, and determining a feature correlation diagram corresponding to the left image and the right image in the binocular stereo image.

For the calculation consideration, the resolution of the feature correlation map is set to 1/4 or 1/8 of the original binocular stereo image, thereby increasing the calculation speed.

Performing feature extraction on the binocular stereo image, and determining a feature correlation diagram corresponding to the left image and the right image in the binocular stereo image, wherein the feature correlation diagram specifically comprises the following steps:

inputting the binocular stereo image into a feature extraction network;

It should be noted that, in order to improve efficiency, the computation of the feature correlation diagram may be implemented by GPU matrix multiplication.

And S106, generating a feature correlation pyramid according to the feature correlation diagram corresponding to the left image and the right image.

And S108, projecting the laser radar point cloud to an image on the appointed side in the binocular stereo image to generate a related laser radar parallax image.

S110, extracting the image on the appointed side in the binocular stereo image, and determining an affinity propagation diagram and context characteristics.

Extracting an image on the appointed side in the binocular stereo image, and determining an affinity propagation diagram and context characteristics, wherein the method specifically comprises the following steps:

And S112, generating a disparity map according to the affinity propagation map and the laser radar disparity image, and searching corresponding related features in the feature related pyramid according to the disparity map.

And S114, inputting the related features and the context features into a GRU updating module, and obtaining a binocular stereo parallax image in an iterative manner.

The GRU update module of the embodiments of the present specification may be provided in plurality.

Inputting the relevant features and the context features into a GRU updating module, and obtaining a binocular stereo parallax image through iterative computation of a plurality of GRU updating modules, wherein the binocular stereo parallax image specifically comprises the following steps:

Similar to 4D correlation volume (correlation volume) constructed in RAFT, in binocular stereo image matching, the embodiment of the specification constructs 3D correlation volume by calculating dot products of left and right image feature maps, and downsamples the last dimension of the 3D correlation volume through a pooling layer, so that the correlation pyramid containing different scales is realized, the correlation volume at each level has different receptive fields, and the correlation volume of the original image resolution is still reserved, so that the construction of the correlation pyramid is completed.

The embodiment of the present specification may further define a search operator, and given the currently estimated disparity, the cost element at the corresponding pixel position may be reversely found in the correlation volume, a 1D mesh is constructed in the correlation volume of each level to define a search range, and then 1D mesh elements of different levels are spliced to form a single feature map.

The loss function of the binocular stereo image parallax matching model is a sparse parallax loss function, a left-right consistent loss function or a smooth loss function.

When the sparse parallax loss function is selected, the method can help to obtain the dense parallax and can supervise parallax estimation; when the left and right consistent loss functions are selected, the estimated left parallax is consistent with the estimated right parallax; and only the appearance and sparse supervision are applied to training, which may cause the estimated parallax to be not smooth and inaccurate, and when a smooth loss function is introduced, the problems can be alleviated.

The embodiments of the present description may train the disparity matching model of the binocular stereo image through a SceneFlow, which is a synthetic image data set including a plurality of binocular stereo images.

The embodiment of the description can solve the binocular stereo matching problem based on an optical flow estimation network RAFT, and the basic idea is still that a correlation volume and multi-level convolution GRU iterative optimization process is constructed by the RAFT, so that global information can be well spread on an image.

Gru (gate recovery unit) is one of Recurrent Neural Networks (RNN). Like LSTM (Long-Short Term Memory), it is proposed to solve the problems of Long-Term Memory and gradient in back-propagation.

Fig. 2 is a schematic structural diagram of a disparity matching model of a binocular stereo image provided in an embodiment of the present specification, where the left side is a right image and a left image of a to-be-processed binocular stereo image, and a lidar disparity image, and the to-be-processed binocular stereo image is subjected to a feature extraction network, and then is subjected to a feature Correlation module (C, Correlation) to generate a feature Correlation pyramid. And (3) after the left image of the binocular stereo image passes through the context extraction network, generating an affinity propagation image and context characteristics. Using the laser radar parallax map d_lAs the initial parallax, and propagating the initial parallax by combining the affinity Propagation map (P, Propagation), obtaining the parallax map d before the first iteration₀Wherein the disparity propagation module is passed through when propagating. Next, the GRU update module will iteratively update the disparity map d, and in the t-th iteration, the following steps will be executed: in a feature-dependent pyramid d_t-1The corresponding relevant features of the parallax position search (L, Lookup) are input to the GRU through the relevant feature query module during search, and also context features output by context extraction and hidden features of the last GRU iteration are input to the GRU (the GRU takes the features output by the feature extraction network as the hidden features during the first iteration), the GRU outputs updated hidden features and outputs a parallax variation delta required to be updated, and the output parallax variation delta and the parallax map d of the previous iteration are output_t-1Adding (+, Addition), and then using the affinity propagation map and the lidar disparity map d_lPerforming disparity Propagation (P) on the added disparity to obtain a disparity map d after the iteration_t. Here, disparity Propagation (P) is also performed iteratively: the first iteration is divided into two steps, firstly, the affinity propagation diagram is used for propagating the parallax diagram once, and then the parallax of the effective laser radar data position is updated to be the laser radar parallax; after M times of transmission, the result of disparity map transmission can be obtained. After the disparity map is updated by the GRU iteration N times, the output disparity is up-sampled (U, Upsampling), and the low-resolution image is up-sampled to the original image resolution, that is, the rightmost image, wherein the up-sampling is performed by an up-sampling module.

Fig. 3 is a schematic structural diagram of a binocular stereoscopic image parallax image acquiring apparatus according to one or more embodiments of the present specification, the apparatus including: an acquisition unit 302, an extraction unit 304, a first generation unit 306, a second generation unit 308, a determination unit 310, a third generation unit 312, and an acquisition unit 314.

The acquiring unit 302 is configured to acquire a binocular stereo image to be processed and a laser radar point cloud, where the binocular stereo image includes a left image and a right image captured by a left camera and a right camera;

the extraction unit 304 is configured to perform feature extraction on the binocular stereo image, and determine a feature correlation diagram corresponding to a left image and a right image in the binocular stereo image;

the first generating unit 306 is configured to generate a feature correlation pyramid according to the feature correlation map corresponding to the left image and the right image;

the second generating unit 308 is configured to project the lidar point cloud to a specified side image in the binocular stereo image, and generate an associated lidar parallax image;

the determining unit 310 is configured to extract an image on a designated side of the binocular stereo image, and determine an affinity propagation map and context features;

the third generating unit 312 is configured to generate a disparity map according to the affinity propagation map and the lidar disparity image, and search for a corresponding relevant feature in the feature-related pyramid according to the disparity map;

the obtaining unit 314 is configured to input the relevant features and the contextual features to the GRU updating module, and obtain a binocular stereo parallax image in an iterative manner.

Fig. 4 is a schematic structural diagram of a binocular stereoscopic image parallax image acquiring apparatus according to one or more embodiments of the present specification, including:

at least one processor; and the number of the first and second groups,

In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually making an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abel (advanced Boolean Expression Language), ahdl (alternate Hardware Description Language), traffic, pl (core universal Programming Language), HDCal (jhdware Description Language), lang, Lola, HDL, laspam, hardward Description Language (vhr Description Language), vhal (Hardware Description Language), and vhigh-Language, which are currently used in most common. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.

The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functions of the various elements may be implemented in the same one or more software and/or hardware implementations of the present description.

As will be appreciated by one skilled in the art, the present specification embodiments may be provided as a method, system, or computer program product. Accordingly, embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.

The description has been presented with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the embodiments of the apparatus, the device, and the nonvolatile computer storage medium, since they are substantially similar to the embodiments of the method, the description is simple, and for the relevant points, reference may be made to the partial description of the embodiments of the method.

The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.

The above description is merely one or more embodiments of the present disclosure and is not intended to limit the present disclosure. Various modifications and alterations to one or more embodiments of the present description will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of one or more embodiments of the present specification should be included in the scope of the claims of the present specification.

Claims

1. A binocular stereo image parallax image acquisition method is characterized by comprising the following steps:

2. The method according to claim 1, wherein the performing feature extraction on the binocular stereo image to determine the feature correlation diagram corresponding to the left image and the right image in the binocular stereo image specifically comprises:

inputting the binocular stereo image into a feature extraction network;

3. The method according to claim 1, wherein the extracting the image at the specified side of the binocular stereo image to determine the affinity propagation map and the contextual features specifically comprises:

4. The method of claim 1, wherein the GRU update module is provided in plurality;

5. The method according to claim 4, wherein the inputting the relevant features and the contextual features into a GRU update module, and obtaining the binocular stereo parallax image through iterative computation by a plurality of GRU update modules specifically comprises:

6. The method according to claim 1, wherein the loss function of the binocular stereo image parallax matching model is one or more of a sparse parallax loss function, a left-right coincidence loss function, and a smooth loss function.

7. The method of claim 1, the resolution of the feature correlation map being 1/4 or 1/8 of the binocular stereo image.

8. A binocular stereoscopic image parallax image acquiring apparatus, comprising:

9. A binocular stereoscopic image parallax image acquiring apparatus, comprising:

at least one processor; and the number of the first and second groups,

10. A non-transitory computer storage medium storing computer-executable instructions, the computer-executable instructions configured to: