Driving assistance image processing method, device, equipment and medium

Abstract

The application discloses a driving assistance image processing method and device, equipment and medium, and relates to the technical field of computers, which comprises the following steps: collecting original images around a target vehicle; determining a target image processing thread from preset image processing threads according to a received target instruction, and determining an image processing node of the target image processing thread; and using the image processing node in the target image processing thread to process the original images correspondingly, so as to obtain a target driving assistance image corresponding to the assistance image acquisition instruction. According to the received target instruction, the target image processing thread and the image processing node in the target image processing thread are determined, that is, unnecessary target image processing threads and image processing nodes are not adopted, and the target image processing thread and the image processing node are flexibly changed, so that the hardware computing power requirement can be reduced and the flexibility of the implementation of the auxiliary driving scene can be improved in the process of obtaining the auxiliary driving image.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to driving assistance image processing methods, devices, equipment, and media. Background Technology

[0002] Advanced driver assistance systems (ADAS) typically utilize fisheye lenses mounted around the vehicle to acquire real-time images. These images require multiple processing steps before they can be used for visual perception by the driving system and display on the vehicle's infotainment system. As a result, the functions of ADAS tend to be more diverse and complex. The image processing system needs to be capable of handling multiple tasks simultaneously and operating efficiently, which places high demands on hardware computing power. Furthermore, the image processing nodes in the current ADAS image acquisition process are fixed, resulting in relatively rigid ADAS scenarios that cannot meet the needs of flexible daily applications.

[0003] In summary, how to reduce hardware computing power requirements and improve the flexibility of implementing assisted driving scenarios in the process of acquiring assisted driving images is a problem that needs to be solved in this field. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a driving assistance image processing method, apparatus, device, and medium that can reduce hardware computing power requirements and improve the flexibility of implementing driving assistance scenarios during the acquisition of driving assistance images. The specific solution is as follows:

[0005] In a first aspect, this application discloses a driving assistance image processing method, including:

[0006] Collect raw images of the area surrounding the target vehicle;

[0007] Based on the received target instruction, the target image processing thread is determined from the preset image processing threads, and the image processing node of the target image processing thread is determined.

[0008] The original image is processed by the image processing node in the target image processing thread to obtain a target driving assistance image corresponding to the target command.

[0009] Optionally, determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, includes:

[0010] If the received target instruction is for acquiring panoramic monitoring images, then the vehicle display thread in the preset image processing thread will be determined as the first target image processing thread.

[0011] The image format conversion node, image color balance node, and image brightness balance node are identified as the image processing nodes of the first target image processing thread.

[0012] Optionally, determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, includes:

[0013] If the target instruction is an instruction for acquiring parking images of fixed parking spaces, then the vehicle display thread and the image recognition thread in the preset image processing thread are respectively determined as the first target image processing thread and the second target image processing thread.

[0014] The image view transformation node and the Tensor transformation node are determined to be the image processing nodes of the second target image processing thread.

[0015] Optionally, determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, includes:

[0016] If the target instruction is an instruction for acquiring remote parking images, then the vehicle display thread, the image recognition thread, and the drivable area detection thread in the preset image processing thread are respectively determined as the first target image processing thread, the second target image processing thread, and the third target image processing thread.

[0017] The image view transformation node, the image stitching node, and the Tensor transformation node are determined to be the image processing nodes of the third target image processing thread.

[0018] Optionally, the step of using the image processing node in the target image processing thread to perform corresponding processing on the original image includes:

[0019] The first target image processing thread is activated based on the buffer buffering method, and the second target image processing thread and the third target image processing thread are activated based on the time-division multiplexing rule;

[0020] The original image is processed using the image processing nodes in the first target image processing thread, the second target image processing thread, and the third target image processing thread.

[0021] Optionally, activating the second target image processing thread and the third target image processing thread based on time-division multiplexing rules includes:

[0022] The first running cycle of the second target image processing thread, the second running cycle of the third target image processing thread, and the acquisition cycle of the original image are determined, and the first actual image processing time of the second target image processing thread and the second actual image processing time of the third target image processing thread are monitored.

[0023] If both the first running cycle and the second running cycle are integer multiples of the acquisition cycle, and the first actual image processing time and the second actual image processing time are less than the first running cycle and the second running cycle, then the second target image processing thread is activated at the current starting point of the acquisition cycle, and the third target image processing thread is activated at the next starting point of the acquisition cycle.

[0024] Optionally, the step of using the image processing node in the target image processing thread to process the original image accordingly to obtain a target driving assistance image corresponding to the target command includes:

[0025] The original image is processed using the image processing node in the target image processing thread to obtain a processed auxiliary image.

[0026] The target offset of the processed auxiliary image in the target offset direction is determined, and the region of interest in the processed auxiliary image is determined based on the target offset direction and the target offset, so that the processed auxiliary image corresponding to the region of interest is determined as the target driving assistance image corresponding to the target command.

[0027] Secondly, this application discloses a driving assistance image processing device, comprising:

[0028] The image acquisition module is used to acquire raw images of the area surrounding the target vehicle.

[0029] The processing node determination module is used to determine the target image processing thread from the preset image processing threads according to the received target instruction, and to determine the image processing node of the target image processing thread;

[0030] The target image acquisition module is used to process the original image using the image processing node in the target image processing thread to obtain a target driving assistance image corresponding to the assistance image acquisition instruction.

[0031] Thirdly, this application discloses an electronic device, including:

[0032] Memory, used to store computer programs;

[0033] A processor is used to execute the computer program to implement the steps of the aforementioned disclosed driving assistance image processing method.

[0034] Fourthly, this application discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the steps of the aforementioned disclosed driving assistance image processing method.

[0035] As can be seen, this application acquires raw images of the area surrounding the target vehicle; determines a target image processing thread from a preset image processing thread based on the received target instruction, and determines the image processing node of the target image processing thread; and uses the image processing node in the target image processing thread to process the raw images accordingly to obtain a target driving assistance image corresponding to the assistance image acquisition instruction. Therefore, by determining the target image processing thread and the image processing node within it based on the received target instruction, unnecessary target image processing threads and image processing nodes are eliminated. Furthermore, the target image processing thread and image processing node can be flexibly changed based on the assistance image acquisition instruction, reducing hardware computing power requirements and increasing the flexibility of implementing assistance driving scenarios during the acquisition of assisted driving images. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0037] Figure 1 This is a flowchart of a driving assistance image processing method disclosed in this application;

[0038] Figure 2 This is a schematic diagram of a specific image processing node disclosed in this application;

[0039] Figure 3 This is a specific thread activation diagram disclosed in this application;

[0040] Figure 4 This is a flowchart of a specific driving assistance image processing method disclosed in this application;

[0041] Figure 5 This is a schematic diagram of a specific initial capture area disclosed in this application;

[0042] Figure 6 This is a schematic diagram of the structure of a driving assistance image processing device disclosed in this application;

[0043] Figure 7 This is a structural diagram of an electronic device disclosed in this application. Detailed Implementation

[0044] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0045] Advanced driver assistance systems (ADAS) typically utilize fisheye lenses mounted around the vehicle to acquire real-time images. These images require multiple processing steps before they can be used for visual perception by the driving system and display on the vehicle's infotainment system. As a result, the functions of ADAS tend to be more diverse and complex. The image processing system needs to be capable of handling multiple tasks simultaneously and operating efficiently, which places high demands on hardware computing power. Furthermore, the image processing nodes in the current ADAS image acquisition process are fixed, resulting in relatively rigid ADAS scenarios that cannot meet the needs of flexible daily applications.

[0046] Therefore, this application provides a driving assistance image processing solution that can reduce hardware computing power requirements and improve the flexibility of implementing driving assistance scenarios during the acquisition of driving assistance images.

[0047] See Figure 1 As shown in the figure, this application discloses a driving assistance image processing method, including:

[0048] Step S11: Acquire raw images of the area surrounding the target vehicle.

[0049] In this embodiment, an image acquisition device pre-installed on the target vehicle is used to acquire raw images of the area around the target vehicle in real time. The image acquisition device can be, for example, a surround-view fisheye camera.

[0050] Step S12: Determine the target image processing thread from the preset image processing threads according to the received target instruction, and determine the image processing node of the target image processing thread.

[0051] It is understandable that there is a mapping relationship between the preset image processing thread and the received target instructions, as shown in the table below:

[0052] Table 1

[0053] Scene function In-vehicle display thread Image recognition thread Detectable driving area thread 360-degree surround view function ON OFF OFF Fixed-point parking space parking function ON ON OFF Remote control parking function ON ON ON

[0054] Each image processing node, when combined, forms a pipeline, which can perform a specific image task. Each pipeline can run independently without blocking others, and a pipeline in the system is represented as an image processing thread. A pre-built mapping relationship between preset image processing threads and target instructions is established. "ON" indicates that the pipeline is open and running in the specified functional scenario, while "OFF" indicates that the pipeline is closed. For example, as shown in Table 1, if a 360-degree surround view function is required (i.e., the target instruction is for acquiring panoramic monitoring images), the vehicle display thread is started, while the image recognition thread and the drivable area detection thread are closed. If a fixed-point parking function is required (i.e., the target instruction is for acquiring fixed-point parking images), the vehicle display thread and the image recognition thread are started, while the drivable area detection thread is closed. If a remote-controlled parking function is required (i.e., the target instruction is for acquiring remote-controlled parking images), the vehicle display thread, the image recognition thread, and the drivable area detection thread are started. By default, all pipelines are in standby mode, meaning threads are not active. Scene functions are triggered by the user, and the state machine sends on / off control signals to each pipeline based on the user's actions. In this embodiment, each thread and image processing node can be flexibly configured, determined according to auxiliary image acquisition instructions, enabling the implementation of various image processing scenarios and providing greater flexibility.

[0055] In a first specific embodiment, determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, specifically includes: if the received target instruction is an instruction for acquiring panoramic monitoring images, then the vehicle display thread in the preset image processing threads is determined as the first target image processing thread; and the image format conversion node, image color balance node, and image brightness balance node are determined as image processing nodes of the first target image processing thread. When the state machine receives the target instruction as an instruction for acquiring panoramic monitoring images, the vehicle display thread in the preset image processing threads is determined as the first target image processing thread, and the image format conversion node, image color balance node, and image brightness balance node are determined as image processing nodes of the first target image processing thread, for example... Figure 2 The diagram shows a specific image processing node. The image format conversion node can convert the original image to YUV format, which can reduce the amount of data storage. When the image needs to be input to the vehicle screen, the YUV format is converted to RGB format. The image color balance node and the image brightness balance node realize the fusion and balance of images from multiple cameras, making the picture input to the vehicle screen more realistic and smooth.

[0056] In a second specific embodiment, determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, specifically includes: if the target instruction is an instruction for acquiring parking images of a fixed parking space, then the vehicle display thread and the image recognition thread in the preset image processing threads are respectively determined as the first target image processing thread and the second target image processing thread; and the image view conversion node and the Tensor conversion node are determined as the image processing nodes of the second target image processing thread. When the state machine receives the target instruction for acquiring parking images of a fixed parking space, the vehicle display thread and the image recognition thread in the preset image processing threads are respectively determined as the first target image processing thread and the second target image processing thread, and the image format conversion node, the image color balance node, and the image brightness balance node are determined as the image processing nodes of the first target image processing thread, and the image view conversion node and the Tensor conversion node are determined as the image processing nodes of the second target image processing thread, for example... Figure 2 As shown, the image view transformation node in the second target image processing thread can transform the fisheye lens image into a columnar distortion-corrected view. The columnar distortion-corrected view can be used for target recognition and lane line recognition, meaning the second target image processing thread can perform target recognition and lane line recognition. The Tensor transformation node is a dedicated input processing node for neural networks, providing image data for deep learning algorithms.

[0057] In a third specific embodiment, determining the target image processing thread from the preset image processing threads based on the received target instruction, and determining the image processing node of the target image processing thread, specifically includes: if the target instruction is an instruction for acquiring remote-controlled parking images, then the vehicle display thread, the image recognition thread, and the drivable area detection thread in the preset image processing threads are respectively determined as the first target image processing thread, the second target image processing thread, and the third target image processing thread; and the image view conversion node, the image stitching node, and the Tensor conversion node are determined as the image processing node of the third target image processing thread. When the state machine receives a target instruction for acquiring remote-controlled parking images, it designates the vehicle display thread, image recognition thread, and drivable area detection thread as the first target image processing thread, the second target image processing thread, and the third target image processing thread, respectively. Furthermore, it designates the image format conversion node, image color balance node, and image brightness balance node as image processing nodes for the first target image processing thread; the image view conversion node and Tensor conversion node as image processing nodes for the second target image processing thread; and the image view conversion node, image stitching node, and Tensor conversion node as image processing nodes for the third target image processing thread. For example... Figure 2 As shown, the image view transformation node can also obtain a top view of the original image. The top view plus the BEV (BirdEye View) image stitching node can realize a panoramic bird's-eye view, which can be used for drivable area search algorithms or ground marking recognition.

[0058] In this embodiment, the step of using the image processing nodes in the target image processing thread to process the original image specifically includes: activating the first target image processing thread based on a buffering method, and activating the second and third target image processing threads based on a time-division multiplexing rule; and using the image processing nodes in the first, second, and third target image processing threads to process the original image accordingly. For example... Figure 3 The diagram illustrates a specific thread activation mechanism. The buffer can store image data from the previous cycle. The cycle of the pipeline displayed on the vehicle's infotainment system is unstable, and the existence of the buffer ensures the real-time performance and independence of the pipeline displayed on the vehicle's infotainment system.

[0059] In this embodiment, activating the second target image processing thread and the third target image processing thread based on time-division multiplexing rules includes: determining the first running cycle of the second target image processing thread, the second running cycle of the third target image processing thread, and the acquisition cycle of the original image; and monitoring the first actual image processing time of the second target image processing thread and the second actual image processing time of the third target image processing thread; if both the first running cycle and the second running cycle are integer multiples of the acquisition cycle, and the first actual image processing time and the second actual image processing time are respectively less than the first running cycle and the second running cycle, then the second target image processing thread is activated at the current starting point of the acquisition cycle, and the third target image processing thread is activated at the next starting point of the acquisition cycle. Figure 3 As shown, different triggering methods are used based on the execution time required for different pipelines. For example, the camera's frame rate is 30fps, meaning one execution cycle is 33ms. The target recognition pipeline's execution cycle is 66ms (15fps) to meet real-time requirements, while the drivable area detection pipeline's execution cycle is 100ms (10fps). Since the actual image processing time for both target recognition and drivable area detection is less than 50ms, the vehicle-mounted display pipeline needs to run continuously, with an execution cycle consistent with the image acquisition cycle of 30fps. If the pipeline's execution cycle is an integer multiple of the image acquisition cycle, such as 33ms*2 = 66ms, 33ms*3 = 99ms, and the actual image processing time is less than the execution cycle, the pipeline is triggered and activated at the instant of the integer multiple of the image acquisition time. If the pipeline's execution cycle is not fixed or the actual image processing time exceeds the image acquisition cycle, a buffer method is used to activate the pipeline. Finally, the actual hardware resource usage time of each pipeline in the system is staggered, thus achieving time-sharing multiplexing. This embodiment of time-sharing multiplexing can reduce the requirements for hardware computing power, avoid the hardware from processing multiple threads at the same time, and reasonably plan the activation time of each thread.

[0060] Step S13: Use the image processing node in the target image processing thread to process the original image accordingly to obtain the target driving assistance image corresponding to the target command.

[0061] By using the image processing nodes in the target image processing thread to process the original image, the processed image can be obtained, which is the target driving assistance image corresponding to the assistance image acquisition command, so as to assist the driver in driving, such as assisting the driver in parking, reversing, etc., and can also assist autonomous driving.

[0062] As can be seen, this application acquires raw images of the area surrounding the target vehicle; determines a target image processing thread from a preset image processing thread based on the received target instruction, and determines the image processing node of the target image processing thread; and uses the image processing node in the target image processing thread to process the raw images accordingly to obtain a target driving assistance image corresponding to the assistance image acquisition instruction. Therefore, by determining the target image processing thread and the image processing node within it based on the received target instruction, unnecessary target image processing threads and image processing nodes are eliminated. Furthermore, the target image processing thread and image processing node can be flexibly changed based on the assistance image acquisition instruction, reducing hardware computing power requirements and increasing the flexibility of implementing assistance driving scenarios during the acquisition of assisted driving images.

[0063] See Figure 4 As shown in the figure, this application discloses a specific driving assistance image processing method, including:

[0064] Step S21: Acquire raw images of the area surrounding the target vehicle.

[0065] Step S22: Determine the target image processing thread from the preset image processing threads according to the received target instruction, and determine the image processing node of the target image processing thread.

[0066] It is understandable that the target instruction can be generated by the user operating the preset display platform, such as the user inputting or selecting the target instruction in the preset display platform; or it can be generated by the user issuing an instruction to the voice system of the target vehicle, such as the user directly speaking the corresponding target instruction inside the target vehicle.

[0067] Step S23: Use the image processing node in the target image processing thread to process the original image accordingly to obtain a processed auxiliary image.

[0068] In this embodiment, the resolution parameters of the target vehicle's driving assistance image can also be obtained. Based on the resolution parameters, it is determined whether to activate the image downsampling node. If activated, the image downsampling node is used to process the original image to obtain a downsampled image. Then, the image processing node in the target image processing thread is used to process the downsampled image accordingly to obtain the target driving assistance image corresponding to the target command. It is understood that if resolution adjustment of the original image is required, the image downsampling node can be used to process the original image, eliminating the need for cameras of various resolutions to obtain original images of different resolutions, thereby reducing the cost of adapting to cameras from different manufacturers.

[0069] Step S24: Determine the target offset of the processed auxiliary image in the target offset direction, and determine the region of interest in the processed auxiliary image based on the target offset direction and the target offset, so as to determine the processed auxiliary image corresponding to the region of interest as the target driving assistance image corresponding to the target command.

[0070] For example Figure 5 The diagram illustrates a specific initial cropping region. An initial cropping region is obtained by initially cropping the image, containing most of the environmental information, such as objects around the vehicle and ground markings. An initial ROI (Region of Interest) is preset, i.e., the initial cropping region is obtained, and offsets in the four directions (up, down, left, and right) are set in the image coordinate system. The initial default value is 0. The actual size of the ROI equals the preset size plus the offsets in the four directions. The image of the default ROI region is downsampled to the size required by the neural network for target recognition and ground lane markings. The targets and ground markings in the ROI are output. The offset of the ROI is adjusted based on the recognized targets and ground markings. Reliable adjustment indicators include the number of recognized targets, the length of the lane lines, and the vehicle's trajectory. For example, when the target vehicle is turning left, the left offset can be increased to expand the left field of view; when the vehicle is going uphill, the upward offset can be increased. When there are too many targets, the offset in target-free areas can be appropriately reduced. When the lane lines are unclear, the up and down offsets can be appropriately increased to increase the perception range. The ROI adjusted in this cycle is re-inputted into the neural network to achieve closed-loop adaptation.

[0071] Therefore, this application can utilize image downsampling nodes to perform resolution transformation on images of any scene to adapt to the requirements of different scenes, without the need to use cameras with different resolutions, thus reducing hardware requirements; furthermore, this application can also plan regions of interest that are more in line with actual conditions, meet the field of vision required by users while driving, and improve the user experience.

[0072] See Figure 6 As shown in the figure, this application discloses a driving assistance image processing device, including:

[0073] Image acquisition module 11 is used to acquire raw images of the area around the target vehicle;

[0074] The processing node determination module 12 is used to determine the target image processing thread from the preset image processing threads according to the received target instruction, and to determine the image processing node of the target image processing thread.

[0075] The target image acquisition module 13 is used to process the original image using the image processing node in the target image processing thread to obtain a target driving assistance image corresponding to the assistance image acquisition instruction.

[0076] As can be seen, this application acquires raw images of the area surrounding the target vehicle; determines a target image processing thread from a preset image processing thread based on the received target instruction, and determines the image processing node of the target image processing thread; and uses the image processing node in the target image processing thread to process the raw images accordingly to obtain a target driving assistance image corresponding to the target instruction. Therefore, by determining the target image processing thread and the image processing node within it based on the received target instruction, unnecessary target image processing threads and image processing nodes are eliminated. Furthermore, the target image processing thread and image processing node can be flexibly changed based on the target instruction, reducing hardware computing power requirements and increasing the flexibility of implementing driving assistance scenarios during the acquisition of driving assistance images.

[0077] Furthermore, embodiments of this application also provide an electronic device. Figure 7 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application.

[0078] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Specifically, it may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the driving assistance image processing method performed by the electronic device disclosed in any of the foregoing embodiments.

[0079] In this embodiment, the power supply 23 is used to provide operating voltage for various hardware devices on the electronic device; the communication interface 24 can create a data transmission channel between the electronic device and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0080] The processor 21 may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor 21 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 21 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 21 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor 21 may also include an AI (Artificial Intelligence) processor, which handles computational operations related to machine learning.

[0081] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored on it include operating system 221, computer program 222 and data 223, etc., and the storage method can be temporary storage or permanent storage.

[0082] The operating system 221 manages and controls the various hardware devices and computer programs 222 on the electronic device to enable the processor 21 to perform calculations and processing on the massive amounts of data 223 in the memory 22. The operating system can be Windows, Unix, Linux, etc. The computer program 222, in addition to including a computer program capable of performing the driving assistance image processing method executed by the electronic device as disclosed in any of the foregoing embodiments, may further include computer programs capable of performing other specific tasks. The data 223 may include data received by the electronic device from external devices, as well as data collected by its own input / output interface 25.

[0083] Furthermore, embodiments of this application also disclose a computer-readable storage medium storing a computer program. When the computer program is loaded and executed by a processor, it implements the method steps performed during the driving assistance image processing process disclosed in any of the foregoing embodiments.

[0084] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0085] The present invention has provided a detailed description of a driving assistance image processing method, apparatus, device, and medium. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A driving assistance image processing method, characterized in that, include: Collect raw images of the area surrounding the target vehicle; Based on the received target instruction, the target image processing thread is determined from the preset image processing threads, and the image processing node of the target image processing thread is determined. The original image is processed by the image processing node in the target image processing thread to obtain a target driving assistance image corresponding to the target instruction. The step of determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, includes: If the target instruction is an instruction for acquiring remote parking images, then the vehicle display thread, image recognition thread, and drivable area detection thread in the preset image processing thread are respectively determined as the first target image processing thread, the second target image processing thread, and the third target image processing thread; the image view conversion node, the image stitching node, and the Tensor conversion node are determined as the image processing nodes of the third target image processing thread.

2. The driving assistance image processing method according to claim 1, characterized in that, The step of determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, further includes: If the received target instruction is for acquiring panoramic monitoring images, then the vehicle display thread in the preset image processing thread will be determined as the first target image processing thread. The image format conversion node, image color balance node, and image brightness balance node are identified as the image processing nodes of the first target image processing thread.

3. The driving assistance image processing method according to claim 2, characterized in that, The step of determining the target image processing thread from the preset image processing threads according to the received target instruction, and determining the image processing node of the target image processing thread, further includes: If the target instruction is an instruction for acquiring parking images of a fixed parking space, then the vehicle display thread and the image recognition thread in the preset image processing thread are respectively determined as the first target image processing thread and the second target image processing thread. The image view transformation node and the Tensor transformation node are determined to be the image processing nodes of the second target image processing thread.

4. The driving assistance image processing method according to claim 3, characterized in that, The step of processing the original image using the image processing node in the target image processing thread includes: The first target image processing thread is activated based on the buffer buffering method, and the second target image processing thread and the third target image processing thread are activated based on the time-division multiplexing rule; The original image is processed using the image processing nodes in the first target image processing thread, the second target image processing thread, and the third target image processing thread.

5. The driving assistance image processing method according to claim 4, characterized in that, The activation of the second target image processing thread and the third target image processing thread based on the time-division multiplexing rule includes: The first running cycle of the second target image processing thread, the second running cycle of the third target image processing thread, and the acquisition cycle of the original image are determined, and the first actual image processing time of the second target image processing thread and the second actual image processing time of the third target image processing thread are monitored. If both the first running cycle and the second running cycle are integer multiples of the acquisition cycle, and the first actual image processing time and the second actual image processing time are less than the first running cycle and the second running cycle, then the second target image processing thread is activated at the current starting point of the acquisition cycle, and the third target image processing thread is activated at the next starting point of the acquisition cycle.

6. The driving assistance image processing method according to any one of claims 1 to 5, characterized in that, The step of processing the original image using the image processing node in the target image processing thread to obtain a target driving assistance image corresponding to the target command includes: The original image is processed using the image processing node in the target image processing thread to obtain a processed auxiliary image. The target offset of the processed auxiliary image in the target offset direction is determined, and the region of interest in the processed auxiliary image is determined based on the target offset direction and the target offset, so that the processed auxiliary image corresponding to the region of interest is determined as the target driving assistance image corresponding to the target command.

7. A driving assistance image processing device, characterized in that, include: The image acquisition module is used to acquire raw images of the area surrounding the target vehicle. The processing node determination module is used to determine the target image processing thread from the preset image processing threads according to the received target instruction, and to determine the image processing node of the target image processing thread; The target image acquisition module is used to process the original image using the image processing node in the target image processing thread to obtain a target driving assistance image corresponding to the assistance image acquisition instruction. The processing node determination module is specifically used for: If the target instruction is an instruction for acquiring remote parking images, then the vehicle display thread, image recognition thread, and drivable area detection thread in the preset image processing thread are respectively determined as the first target image processing thread, the second target image processing thread, and the third target image processing thread; the image view conversion node, the image stitching node, and the Tensor conversion node are determined as the image processing nodes of the third target image processing thread.

8. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the driving assistance image processing method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, Used to store a computer program; wherein, when the computer program is executed by a processor, it implements the steps of the driving assistance image processing method as described in any one of claims 1 to 6.

Images