Image processing method and reverse driving auxiliary driving system

Abstract

This application discloses an image processing method for a reversing assist driving system. The method acquires initial images from at least two cameras; then, it obtains feature points and corresponding target coordinates from the initial images; next, it determines the overlapping area between the initial images based on the feature points and the target coordinates; finally, it stitches the initial images together based on the overlapping area to obtain a wide-angle image. This application stitches together the initial images from at least two cameras to obtain a wide-angle image with a large display range, facilitating the driver's understanding of the environment behind the vehicle, assisting the driver in reversing, and improving reversing safety.

Description

Technical Field

[0001] This application belongs to the field of vehicle reversing image processing technology research, and specifically relates to an image processing method and a reversing auxiliary driving system. Background Technology

[0002] Large special vehicles are significantly wider and taller than regular cars, making reversing much more difficult and dangerous. Existing reversing assistance systems have a limited field of vision, failing to display the entire environment behind the vehicle, which can easily create blind spots and lead to accidents. Therefore, existing reversing assistance systems are not suitable for use on large special vehicles.

[0003] In view of the prior art, this application proposes an image processing method to solve the problem of small visual range of image display in existing reversing driver assistance systems. Summary of the Invention

[0004] To address the shortcomings of the prior art, this application provides an image processing method for a reversing assist driving system. The method acquires initial images from at least two cameras; then, it obtains feature points and corresponding target coordinates from the initial images; next, it determines the overlapping area between the initial images based on the feature points and the target coordinates; finally, it stitches the initial images together based on the overlapping area to obtain a wide-angle image. This application stitches together the initial images from at least two cameras to obtain a wide-angle image with a large display range, facilitating the driver's understanding of the environment behind the vehicle, assisting the driver in reversing, and improving reversing safety.

[0005] In a first aspect, this application provides an image processing method for a reversing assist driving system, wherein the vehicle includes a camera assembly comprising at least two cameras, characterized in that the method comprises:

[0006] Acquire initial images from at least two of the cameras;

[0007] Based on the initial image, feature points and the target coordinates corresponding to the feature points are obtained;

[0008] Based on the feature points and the target coordinates, determine the overlapping region between the initial images;

[0009] The initial image is stitched together based on the overlapping area to obtain a wide-angle image.

[0010] Optionally, obtaining the feature points and the target coordinates corresponding to the feature points based on the initial image includes:

[0011] Obtain the coordinate system of each of the initial images;

[0012] Based on each initial image and the coordinate system, determine the common feature points between each initial image and the target coordinates corresponding to the feature points; wherein, the feature points are objects or markers common to at least two initial images.

[0013] Optionally, determining the common feature points and target coordinates corresponding to the feature points among the initial images based on the initial images and the coordinate system includes:

[0014] Acquire the initial images continuously captured by each of the cameras and the acquisition order of the initial images;

[0015] Based on the initial image, the acquisition order, and the coordinate system, determine the common feature points and the target coordinates corresponding to the feature points among each initial image.

[0016] Optionally, determining the common feature points and the target coordinates corresponding to the feature points among each of the initial images based on the initial images, the acquisition order, and the coordinate system includes:

[0017] A key initial image is obtained based on the initial image and the acquisition order; wherein, the key initial image is the initial image whose acquisition order is a first preset value;

[0018] The feature points are obtained based on the key initial image;

[0019] The target coordinates corresponding to the feature points are obtained based on the initial image, the feature points, and the coordinate system; wherein, the target coordinates are obtained by performing a weighted average calculation on the feature point coordinates of the initial image whose acquisition order is from the first preset value to the second preset value.

[0020] Optionally, determining the overlapping region between the initial images based on the feature points and the target coordinates includes:

[0021] The boundary feature points of the initial image are obtained based on the feature points and the target coordinates;

[0022] Obtain the boundary coordinates based on the boundary feature points;

[0023] The overlapping region is obtained based on the boundary coordinates, wherein the boundary coordinates in the same initial image are connected to obtain a boundary line, and the boundary line divides the initial image into a non-overlapping region and an overlapping region.

[0024] Secondly, this application proposes a reversing assist driving system for assisting a vehicle in reversing, characterized in that the reversing image is processed using the image processing method described in the first aspect, and the reversing assist driving system includes a camera assembly, an image processor, a main controller, and a display.

[0025] The camera assembly includes a first camera, a second camera, and a third camera. The first camera and the second camera are respectively located on both sides of the upper end of the rear of the vehicle, and the third camera is located at the lower end of the rear of the vehicle. The first camera, the second camera, and the third camera are respectively connected to the image processor.

[0026] The image processor includes a processing chip;

[0027] The main controller includes a CPU processor, which is connected to the image processor.

[0028] The display is connected to the image processor.

[0029] Furthermore, the reversing assist driving system includes a radar component, which includes multiple radar probes and a radar controller. The multiple radar probes are spaced apart at the rear end of the vehicle, and the multiple radar probes are connected to the radar controller, which is connected to the display.

[0030] Thirdly, this application provides an image processing apparatus, the image processing apparatus comprising:

[0031] The acquisition module is used to acquire initial images from at least two cameras;

[0032] The first determining module obtains feature points and target coordinates corresponding to the feature points based on the initial image;

[0033] The second determining module determines the overlapping region between the initial images based on the feature points and the target coordinates;

[0034] The third determining module stitches the initial image according to the overlapping area to obtain a wide-angle image.

[0035] Fourthly, this application proposes a readable medium comprising execution instructions, wherein when a processor of an electronic device executes the execution instructions, the electronic device performs the image processing method as described in the first aspect.

[0036] Fifthly, this application proposes an electronic device comprising a processor and a memory storing execution instructions, wherein when the processor executes the execution instructions stored in the memory, the processor performs the image processing method as described in the first aspect.

[0037] Compared with existing technologies, the beneficial effects of this application are as follows: This application provides an image processing method for a reversing assist driving system. The method acquires initial images captured by at least two cameras; then, it obtains feature points and corresponding target coordinates based on the initial images; next, it determines the overlapping area between the initial images based on the feature points and the target coordinates; and then, it stitches the initial images together based on the overlapping area to obtain a wide-angle image. This application stitches together the initial images captured by at least two cameras to obtain a wide-angle image with a large display range, making it easier for the driver to understand the environment behind the vehicle, assisting the driver in reversing, and improving reversing safety. Attached Figure Description

[0038] To more clearly illustrate the embodiments of this application or the existing technical solutions, 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 some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of the image processing method described in one embodiment of this application;

[0040] Figure 2 This is a schematic diagram of the reversing assist driving system described in one embodiment of this application;

[0041] Figure 3 This is a schematic diagram of the image processing apparatus described in one embodiment of this application;

[0042] Figure 4 This is a schematic diagram of the structure of the electronic device described in one embodiment of this application. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0044] Large special-purpose vehicles are difficult to reverse due to their large size, requiring reversing assistance systems to reduce the probability of accidents. However, existing reversing assistance systems are designed for regular cars, and when applied to large special-purpose vehicles, the displayed image is too narrow to show the entire environment behind the vehicle, creating blind spots. This makes large special-purpose vehicles equally prone to accidents when reversing. To address these issues, this application proposes an image processing method for a reversing assistance system. The method acquires initial images from at least two cameras; then, it obtains feature points and corresponding target coordinates from the initial images; next, it determines the overlapping area between the initial images based on the feature points and target coordinates; finally, it stitches the initial images together based on the overlapping area to obtain a wide-angle image. This application stitches together the initial images from at least two cameras to obtain a wide-angle image with a large display range, facilitating the driver's understanding of the environment behind the vehicle, assisting the driver in reversing, and improving reversing safety.

[0045] The various non-limiting embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0046] See appendix Figure 1 This paper illustrates an image processing method according to an embodiment of the present application, which is used in a reversing assist driving system. In this embodiment, the method may include, for example, the following steps:

[0047] S101: Acquire initial images captured by at least two of the cameras.

[0048] A single camera captures an image with a small display area, which cannot show the entire environment behind the vehicle. Installing at least two cameras at the rear of the vehicle allows for a complete view of the environment. Acquiring initial images from at least two cameras provides a wider field of view, yielding more information about the environment behind the vehicle. This better assists the driver in reversing and reduces the probability of reversing accidents.

[0049] S102: Obtain feature points and target coordinates corresponding to the feature points based on the initial image.

[0050] Each camera obtains a different initial image, and some environmental images are the same in the initial images obtained by different cameras. To stitch together the initial images obtained by at least two cameras, it is necessary to find the overlapping parts of the initial images and overlap them to stitch together multiple initial images to display the environment behind the vehicle. To distinguish the overlapping areas of at least two initial images, feature extraction is required for each initial image to obtain objects or markers common to at least two initial images as feature points. To better stitch together the at least two initial images later, the target coordinates corresponding to the feature points can be obtained simultaneously with obtaining the feature points. The target coordinates are the coordinates of the feature points on the initial images.

[0051] In this embodiment, by obtaining feature points and their corresponding target coordinates from the initial images, the coordinate system of each initial image can be obtained. Then, based on each initial image and the coordinate system, common feature points and their corresponding target coordinates are determined among the initial images. The feature points are objects or markers common to at least two initial images. Each initial image has a coordinate system; by identifying the common feature points of each initial image and initial images acquired by other different cameras, the target coordinates of the feature points in the coordinate system of the initial image are obtained.

[0052] Furthermore, by determining the common feature points and corresponding target coordinates of each initial image based on the initial image and the coordinate system, the initial images continuously acquired by each camera and the acquisition order of the initial images can be obtained. Then, based on the initial images, the acquisition order, and the coordinate system, the common feature points and corresponding target coordinates of each initial image are determined. The initial images acquired by the cameras are frame by frame, meaning each camera continuously acquires multiple frames of the initial images. The initial images acquired by multiple cameras at the same time need to correspond according to the acquisition order. The initial images acquired by multiple cameras at the same time can be stitched together, so that the stitched image can accurately display the environment behind the vehicle at that point in time. If the initial images acquired by different cameras at different times are stitched together, the stitched image will be distorted, and the subsequent algorithm required to correct the distorted image is large. Because the vehicle is constantly moving, the initial images acquired by multiple different cameras at the same time are corresponding to each other, so the stitched image will not be distorted.

[0053] In this embodiment, based on the initial image, the acquisition order, and the coordinate system, the common feature points and their corresponding target coordinates are determined among each initial image. A key initial image can be obtained based on the initial image and the acquisition order; wherein the key initial image is the initial image whose acquisition order is a first preset value. Then, the feature points are obtained based on the key initial image; then, the target coordinates corresponding to the feature points are obtained based on the initial image, the feature points, and the coordinate system. The target coordinates are obtained by performing a weighted average calculation on the feature point coordinates of the initial images acquired in acquisition orders from the first preset value to the second preset value. If each frame of the initial image acquired by the camera is stitched together, the number of initial images to be stitched would be too large, slowing down the stitching process and preventing timely output of the stitched initial images. Therefore, the initial image acquired in the first preset value can be used as the key initial image, and other initial images can be used as ordinary initial images. The process involves acquiring key initial images from different cameras, identifying feature points within these images, and then obtaining the coordinates of these feature points in each initial image acquired by the camera. The coordinates of the feature points in all initial images acquired in sequences from the first preset value to the second preset value are also obtained. A weighted average of these feature point coordinates is then calculated to obtain the target coordinates. Using these target coordinates as the coordinates of the feature points in the key initial images allows for the stitching of only the key initial images acquired by multiple cameras to obtain a wide-angle image for display. This reduces the number of images to be stitched, decreases the workload, and speeds up the initial image stitching process.

[0054] The acquisition of feature points is crucial for image registration technology. Accurate feature points ensure successful feature matching. However, to guarantee the real-time performance of the system, the initial images are collected in a period of 40 frames, with the 1st, 40th, 80th...40Nth frames being key initial images, and the others being ordinary initial images, where N is a non-zero natural number. Feature point extraction and registration are performed on the key initial images, while the ordinary initial images use the feature points from the previous key initial image for image registration. The target coordinates of the feature points are the weighted average of the coordinates of the feature points in the ordinary initial image and the previous key initial image. That is, in the images captured by the same camera, the coordinates of a feature point are (3, 4) in the first key initial image, (3.2, 4.1) in the second ordinary initial image, (3.3, 4.2) in the third ordinary initial image, ... and (3.2, 4.1) in the Nth ordinary initial image. Then the X-axis of the target coordinates is (3 + 3.2 + 3.3 + ... + 3.2) / N, and the Y-axis of the target coordinates is (4 + 4.1 + 4.2 + ... + 4.1) / N. The key initial images are all initial images with the acquisition sequence of 40M+1, where M is a natural number and M can be zero.

[0055] S103: Determine the overlapping area between the initial images based on the feature points and the target coordinates.

[0056] Based on the feature points and the target coordinates, the overlapping areas of the initial images captured by different cameras can be clearly identified. The overlapping areas of each initial image can be divided according to the feature points and the target coordinates, which can improve the realism of the image after stitching the initial images and reduce the distortion of the stitched image.

[0057] In this embodiment, determining the overlapping region between the initial images based on the feature points and the target coordinates involves obtaining the boundary feature points of the initial images based on the feature points and the target coordinates; then, obtaining the boundary coordinates based on the boundary feature points; and finally, obtaining the overlapping region based on the boundary coordinates. Connecting the boundary coordinates within the same initial image yields a boundary line, which divides the initial image into non-overlapping and overlapping regions. The boundary feature points are those that clearly form boundaries; that is, one side of the region has more feature points, while the other side has none. Connecting the boundary coordinates divides the initial image into non-overlapping and overlapping regions. The overlapping region is a graphical area shared by at least two initial images. Overlapping these regions yields a wide-angle image with a broader field of view.

[0058] S104: The initial image is stitched together according to the overlapping area to obtain a wide-angle image.

[0059] To obtain a complete wide-angle image, the overlapping areas of at least two initial images are overlapped to avoid the appearance of two identical objects or markers in the wide-angle image, thus preventing image duplication and improving the accuracy of image display while increasing the visual angle.

[0060] As attached Figure 2 As shown in the illustration, in a specific embodiment of the reversing assist driving system described in this application, the reversing assist driving system processes reversing images using the image processing method described above. The reversing assist driving system includes a camera assembly, an image processor, a main controller, and a display. The camera assembly includes a first camera, a second camera, and a third camera. The first and second cameras are respectively located on either side of the upper rear of the vehicle, and the third camera is located at the lower rear of the vehicle. The first, second, and third cameras are each connected to the image processor. The image processor includes a processing chip. The main controller includes a CPU processor, which is connected to the image processor. The display is connected to the image processor. The first, second, and third cameras of the camera assembly capture environmental images behind the vehicle and transmit the initial images to the image processor. The image processor stitches the initial images together to obtain a wide-angle image with a broad field of view. The main controller controls the image processor to process the initial image. In addition, the main controller controls the image processor to receive the initial image captured by the camera component. After the image processor stitches the initial image, it controls the image processor to send the wide-angle image to the display. The wide-angle image is displayed on the display, which can be set in front of the driver's cab, on one side of the steering wheel, for easy viewing by the driver.

[0061] In this embodiment, the image processor is mounted on an XC158 series waterproof component fixed to a circuit board. The image processor's housing meets IP65 waterproof requirements. The XC158 component and the internal components are fixedly soldered to the circuit board using cable ties. The processing chip uses an Insetri industrial-grade TW6865 for image decoding and conversion. The TW6865 and the processing chip exchange data via PCIe. Wide-angle image output uses an ADI ADV7393 chip for image encoding, resulting in high overall reliability. The image processor's power supply front end uses a TI industrial-grade DC-DC isolation module PTB78560CAH, effectively preventing interference from external cameras or other sources. The core power supply and the power supplies for other peripherals all use TI industrial-grade chips, and the analog power supply is entirely isolated and powered separately using ferrite beads, ensuring stable operation of the entire system. The CPU processor uses a standard Freescale i.MX6Q quad-core processor with an ARM Cortex-A9 core and a clock speed of up to 1GHz. It is compatible with single-core and dual-core processors and can realize multi-screen asynchronous output display, dual-screen display playback of 720P video, and other functions. The i.MX6Q supports floating-point and multi-core operations. In particular, for multi-core operations, the i.MX6Q starts up with all four cores upon power-up with Freescale's BSP package, and the four cores work in a balanced manner, which significantly improves the computing efficiency.

[0062] The image processor can also perform perspective transformation and deblurring on the initial image. Perspective transformation projects the image onto another plane to form an image. Since the camera is not mounted perpendicular to the ground, a bird's-eye view transformation is needed to convert the images from each camera channel into a top-down view for subsequent algorithm processing. Because the camera is mounted on a moving vehicle, the image data captured by the camera will have motion blur due to the instantaneous and rapid relative motion / shaking during the vehicle's movement. To ensure the accuracy of panoramic image registration, the image processor performs deblurring on the images from each channel to restore image details.

[0063] The image processor acquires the initial images captured by the first and second cameras, and performs intrinsic parameter correction to reduce image distortion caused by lens distortion. It then stitches and merges the two images into a single, smoothly transitioning, wide-angle image. This wide-angle image provides a relatively high field of view of the rear of the vehicle, helping the driver avoid collisions between the rear of the vehicle or critical vehicle components and objects. The image processor also acquires the initial image captured by the third camera, performs intrinsic parameter correction to reduce image distortion caused by lens distortion, and then outputs the image. This initial image from the third camera provides a relatively low field of view of the rear of the vehicle, compensating for the blind spots in the lower middle part of the rear of the vehicle in the wide-angle image, helping the driver to see the ground scene and avoid collision hazards.

[0064] Furthermore, the reversing assist driving system includes a radar component, which comprises multiple radar probes and a radar controller. The multiple radar probes are spaced apart at the rear end of the vehicle, and are connected to the radar controller, which is connected to the display. The radar probes are used to sense the distance between the rear of the vehicle and other objects. When the distance is less than a preset distance, an alarm or alert sound can be sent. The radar controller acquires the distance sensed by the radar probes and sends the distance to the display for display. By positioning multiple radars at different locations at the rear of the vehicle, the system can sense the distance to multiple objects at various points on the rear of the vehicle, preventing collisions in multiple areas and improving safety.

[0065] The main functions of the radar controller include powering the radar probe, converting the distance information collected by the radar probe into valid values, and converting it into CAN protocol data for output to the CAN bus. The radar controller adopts a sealed design, with waterproof conductive rubber strips filling the space between the upper and lower covers, and secured by 12 triple-screws around the perimeter. It uses seven M12 aviation connectors (IP67) for external connections, achieving an overall protection rating of IP65. The radar controller consists of a filter, power module, circuit board, radar controller, and connectors. The CPU uses the Infineon XC2000 series automotive-grade microcontroller XC2287M with a main frequency of 80MHz. The radar controller hardware adopts a modular design, which can be divided into a power management module, radar control module, control module, and interface unit. The power management module, radar control module, and interface unit use standard modules. The control module is the core control module of the system, and a highly integrated controller is selected during the design process to manage and control various interfaces and complete the core functions of the system. The circuit principle adopts a modular design, and the printed circuit board adopts an integrated design. After the radar controller is connected to the chassis CAN power supply system, it supplies power to the reversing radar probes through an external interface. Once the chassis CAN system is powered on, the ultrasonic radar controller powers on first, followed by the radar probes. After powering on, the radar probes collect distance information of obstacles within a certain range behind the vehicle in real time and provide it to the radar controller for processing. After powering on, the radar controller first initializes, turns on the radar controller and power output switches, and then waits for and continuously processes the distance information uploaded by the radar probes. After conversion, it uploads the information to the display at a fixed 100ms interval.

[0066] After the radar controller starts working, it sends a transmission signal to the radar probe. Upon receiving the signal, the radar probe emits ultrasonic waves. When an obstacle is detected within its effective range, the ultrasonic waves impact and return. The radar probe receives the reflected wave and sends an echo signal to the radar controller. The radar controller calculates the distance to the obstacle based on the time interval between the transmission and echo signals and converts the distance data into CAN data. The central control software parses the radar CAN data, calculates the actual radar distance, filters the results, and displays radar stripes of corresponding colors based on the distance. The driver can judge the distance to the obstacle by viewing the radar stripe colors on the display.

[0067] As attached Figure 3 The image shown is a specific embodiment of the image processing apparatus of this application, the image processing apparatus comprising:

[0068] The acquisition module is used to acquire initial images from at least two cameras;

[0069] The first determining module obtains feature points and target coordinates corresponding to the feature points based on the initial image;

[0070] The second determining module determines the overlapping region between the initial images based on the feature points and the target coordinates;

[0071] The third determining module stitches the initial image according to the overlapping area to obtain a wide-angle image.

[0072] Optionally, the first determining module is used to:

[0073] Obtain the coordinate system of each of the initial images;

[0074] Based on each initial image and the coordinate system, determine the common feature points between each initial image and the target coordinates corresponding to the feature points; wherein, the feature points are objects or markers common to at least two initial images.

[0075] Optionally, the first determining module is used to:

[0076] Acquire the initial images continuously captured by each of the cameras and the acquisition order of the initial images;

[0077] Based on the initial image, the acquisition order, and the coordinate system, determine the common feature points and the target coordinates corresponding to the feature points among each initial image.

[0078] Optionally, the first determining module is used to:

[0079] A key initial image is obtained based on the initial image and the acquisition order; wherein, the key initial image is the initial image whose acquisition order is a first preset value;

[0080] The feature points are obtained based on the key initial image;

[0081] The target coordinates corresponding to the feature points are obtained based on the initial image, the feature points, and the coordinate system; wherein, the target coordinates are obtained by performing a weighted average calculation on the feature point coordinates of the initial image whose acquisition order is from the first preset value to the second preset value.

[0082] Optionally, the second determining module is used to:

[0083] The boundary feature points of the initial image are obtained based on the feature points and the target coordinates;

[0084] Obtain the boundary coordinates based on the boundary feature points;

[0085] The overlapping region is obtained based on the boundary coordinates, wherein the boundary coordinates in the same initial image are connected to obtain a boundary line, and the boundary line divides the initial image into a non-overlapping region and an overlapping region.

[0086] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include RAM, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk storage device. Of course, the electronic device may also include other hardware required for other services.

[0087] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0088] Memory is used to store instructions for execution. Specifically, instructions for execution are computer programs that can be executed. Memory can include main memory and non-volatile memory, and it provides the processor with execution instructions and data.

[0089] In one possible implementation, the processor reads the corresponding execution instructions from non-volatile memory into memory and then executes them. Alternatively, it may obtain the corresponding execution instructions from other devices to form an image processing method at the logical level. The processor executes the execution instructions stored in memory to implement the image processing method provided in any embodiment of this application through the executed instructions.

[0090] The above is as stated in this application. Figure 1 The image processing method provided in the illustrated embodiment can be applied to or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by software instructions. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor.

[0091] The steps of the method disclosed in the embodiments of this application can be directly manifested as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0092] This application also proposes a readable medium that stores execution instructions. When the stored execution instructions are executed by the processor of an electronic device, the electronic device can execute the image processing method provided in any embodiment of this application, and specifically perform the above-mentioned image processing method.

[0093] The electronic devices described in the foregoing embodiments may be computers.

[0094] Those skilled in the art will understand that the embodiments of this application can be provided as methods or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or a combination of software and hardware.

[0095] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0096] It should also be noted that 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 limitation, 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.

[0097] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An image processing method for a reversing assist driving system, wherein the vehicle includes a camera assembly comprising at least two cameras, characterized in that, The method includes: Acquire initial images from at least two of the cameras; Based on the initial image, feature points and the target coordinates corresponding to the feature points are obtained; Based on the feature points and the target coordinates, determine the overlapping region between the initial images; The initial image is stitched together based on the overlapping area to obtain a wide-angle image; The step of obtaining feature points and corresponding target coordinates based on the initial image includes: Obtain the coordinate system of each of the initial images; Based on each initial image and the coordinate system, determine the common feature points between each initial image and the target coordinates corresponding to the feature points; wherein, the feature points are objects or markers common to at least two initial images; The step of determining the common feature points and the target coordinates corresponding to the feature points among the initial images based on the initial images and the coordinate system includes: Acquire the initial images continuously captured by each of the cameras and the acquisition order of the initial images; Based on the initial image, the acquisition order, and the coordinate system, determine the common feature points and the target coordinates corresponding to the feature points among each initial image; The method further includes: The initial image is subjected to image perspective transformation and deblurring processing; the image perspective transformation is to transform the images captured by each camera into a bird's-eye view and convert them into a top view; the deblurring processing is to restore the image blur caused by vehicle movement to improve the accuracy of subsequent image registration. The step of determining the overlapping region based on the feature points and the target coordinates further includes: Connect the boundary coordinates in the same initial image to form a boundary line, which divides the initial image into non-overlapping and overlapping regions; The feature point extraction and registration are performed only on the key initial image; the feature points of the previous key initial image are used for registration of ordinary initial images. The key initial image is an initial image acquired in sequence 40N+1, where N is a natural number; the target coordinates of the feature points are the weighted average of the coordinates of the feature points in the ordinary initial image and the previous key initial image.

2. The image processing method as described in claim 1, characterized in that, Determining the overlapping region between the initial images based on the feature points and the target coordinates includes: The boundary feature points of the initial image are obtained based on the feature points and the target coordinates; Obtain the boundary coordinates based on the boundary feature points; The overlapping region is obtained based on the boundary coordinates.

3. A reversing assist driving system for assisting a vehicle in reversing, characterized in that, The reversing image is processed using the image processing method described in any one of claims 1-2, and the reversing assist driving system includes a camera assembly, an image processor, a main controller, and a display; The camera assembly includes a first camera, a second camera, and a third camera. The first camera and the second camera are respectively located on both sides of the upper end of the rear of the vehicle, and the third camera is located at the lower end of the rear of the vehicle. The first camera, the second camera, and the third camera are respectively connected to the image processor. The image processor includes a processing chip; The main controller includes a CPU processor, which is connected to the image processor. The display is connected to the image processor.

4. The reversing assist driving system as described in claim 3, characterized in that, The reversing assist driving system includes a radar component, which includes multiple radar probes and a radar controller. The multiple radar probes are spaced apart at the rear end of the vehicle and are connected to the radar controller, which is connected to the display.

5. An image processing apparatus, characterized in that, The image processing device includes: The acquisition module is used to acquire initial images from at least two cameras; The first determining module obtains feature points and target coordinates corresponding to the feature points based on the initial image; The second determining module determines the overlapping region between the initial images based on the feature points and the target coordinates; The third determining module stitches the initial image according to the overlapping area to obtain a wide-angle image; The step of obtaining feature points and corresponding target coordinates based on the initial image includes: Obtain the coordinate system of each of the initial images; Based on each initial image and the coordinate system, determine the common feature points between each initial image and the target coordinates corresponding to the feature points; wherein, the feature points are objects or markers common to at least two initial images; The step of determining the common feature points and the target coordinates corresponding to the feature points among the initial images based on the initial images and the coordinate system includes: Acquire the initial images continuously captured by each of the cameras and the acquisition order of the initial images; Based on the initial image, the acquisition order, and the coordinate system, determine the common feature points and the target coordinates corresponding to the feature points among each initial image; The device is also used for: The initial image is subjected to image perspective transformation and deblurring processing; the image perspective transformation is to transform the images captured by each camera into a bird's-eye view and convert them into a top view; the deblurring processing is to restore the image blur caused by vehicle movement to improve the accuracy of subsequent image registration. The step of determining the overlapping region based on the feature points and the target coordinates further includes: Connect the boundary coordinates in the same initial image to form a boundary line, which divides the initial image into non-overlapping and overlapping regions; The feature point extraction and registration are performed only on the key initial image; the feature points of the previous key initial image are used for registration of ordinary initial images. The key initial image is an initial image acquired in sequence 40N+1, where N is a natural number; the target coordinates of the feature points are the weighted average of the coordinates of the feature points in the ordinary initial image and the previous key initial image.

6. A readable medium, characterized in that, The readable medium includes execution instructions that, when executed by the processor of the electronic device, enable the electronic device to perform the method as described in any one of claims 1-2.

7. An electronic device, characterized in that, The electronic device includes a processor and a memory storing execution instructions. When the processor executes the execution instructions stored in the memory, the processor performs the method as described in any one of claims 1-2.

Images