Reachable space detection method, device, vehicle-mounted terminal and storage medium

Abstract

The present application discloses a method, device, vehicle-mounted terminal and storage medium for detecting accessible space, and belongs to the field of data processing technology. Applied to a vehicle-mounted terminal, the method includes: obtaining a first fisheye image through a fisheye camera; obtaining a first environment image through a shooting camera; distorting the first environment image to obtain a second fisheye image; inputting the first fisheye image and the second fisheye image into a space detection model, and detecting the accessible space of the vehicle's surrounding environment through the space detection model. The present application captures a first environment image through a shooting camera, and distorts the first environment image to obtain a second fisheye image, and inputs both the first fisheye image and the second fisheye image into the space detection model, and detects the accessible space of the vehicle's surrounding environment through the space detection model, so that the image data used by the vehicle-mounted terminal for the detection of the accessible space is more accurate, thereby improving the accuracy of the vehicle-mounted terminal's detection of the accessible space.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, device, vehicle-mounted terminal and storage medium for detecting an accessible space. Background Art

[0002] With the rapid development of science and technology, various types of vehicles have appeared in people's daily lives, and the use of on-board terminals in vehicles has become more and more common. For example, users can use on-board terminals to perform space detection, play audio, and locate positions.

[0003] In terms of space detection, different cameras are usually used to collect environmental images of the surrounding environment, and based on the environmental images, the accessible space in the surrounding environment is detected. For example, the vehicle can control the fisheye camera through the vehicle terminal to collect environmental images of the surrounding environment, obtain fisheye images, use the fisheye images to train the neural network model for detecting the accessible space, obtain the required neural network model, and use the neural network model to detect the accessible space. Among them, when the fisheye image obtained by the vehicle through the fisheye camera has fewer features, it will affect the results of the accessible space detection, resulting in low accuracy of the on-board terminal in detecting the accessible space. Summary of the invention

[0004] The embodiments of the present application provide a method, device, vehicle-mounted terminal and storage medium for detecting a reachable space, which can increase the image data used in the detection of the reachable space and improve the accuracy of the detection of the reachable space.

[0005] In one aspect, an embodiment of the present application provides a method for detecting a reachable space, which is applied to a vehicle-mounted terminal, and the method includes:

[0006] Acquire a first fisheye image by using a fisheye camera;

[0007] Acquire a first environment image by shooting a camera;

[0008] Distorting the first environment image to obtain a second fisheye image;

[0009] The first fisheye image and the second fisheye image are input into a space detection model, and the accessible space of the vehicle surrounding environment is detected by the space detection model.

[0010] Optionally, inputting the first fisheye image and the second fisheye image into a space detection model, and detecting the accessible space of the surrounding environment of the vehicle by using the space detection model includes:

[0011] Performing feature fusion on the first fisheye image and the second fisheye image to obtain a target fused image;

[0012] According to the target fusion image, the accessible space of the vehicle surrounding environment is detected by the space detection model.

[0013] Optionally, spatially fusing the first fisheye image and the second fisheye image to obtain a target fused image includes:

[0014] Performing image recognition on the first fisheye image by using the spatial detection model to obtain a first image feature;

[0015] Performing image recognition on the second fisheye image by using the spatial detection model to obtain second image features;

[0016] The first fisheye image and the second fisheye image are fused according to the first image feature and the second image feature to obtain the target fused image.

[0017] Optionally, detecting the accessible space of the vehicle surrounding environment by using the space detection model according to the target fusion image includes:

[0018] Extracting features from the target fused image to obtain a feature pyramid, wherein the feature pyramid includes multiple layers of image features of the target fused image;

[0019] According to the image features of the target layer in the feature pyramid and the regression model, an initial boundary of the reachable space in the target fused image is obtained, wherein the target layer is a layer in the feature pyramid containing the largest number of features, and the regression model is a model trained based on at least two sample layers and obtaining an initial boundary with the same width as the sample layer;

[0020] According to the initial boundary, the accessible space of the vehicle surrounding environment is detected by the space detection model.

[0021] Optionally, after detecting the reachable space of the vehicle surrounding environment according to the target fused image, the method further includes:

[0022] According to the target fused image, an attribute feature image is acquired, wherein the attribute feature image includes attribute features of each obstacle in the target fused image;

[0023] Performing attribute classification on the attribute feature image to obtain an attribute graph corresponding to each obstacle attribute, wherein the attribute graph includes a confidence level of the obstacle attribute corresponding to each pixel point in the target fused image;

[0024] According to the attribute map, an obstacle attribute of an obstacle at a first position in the target fused image is determined, where the first position is a pixel point position included in any obstacle in the target fused image.

[0025] Optionally, distorting the first environment image to obtain a second fisheye image includes:

[0026] Acquire a first distorted image according to the first environment image and a first distortion function;

[0027] Performing coordinate screening on the first distorted image to determine a target area image in the first distorted image;

[0028] The second fisheye image is acquired according to the target area image.

[0029] Optionally, acquiring the second fisheye image according to the target area image includes:

[0030] Performing magnification and edge trimming processing on the target area image to obtain a target fisheye image;

[0031] Acquire a first added image according to the first fisheye image, where the first added image is a blind area image in the first fisheye image;

[0032] The first added image is added to the target fisheye image to generate the second fisheye image. In another aspect, an embodiment of the present application provides a detection device for a reachable space, the device is applied to a vehicle-mounted terminal, and the device includes:

[0033] A first acquisition module, used for acquiring a first fisheye image through a fisheye camera;

[0034] A second acquisition module, used for acquiring a first environment image through a shooting camera;

[0035] A third acquisition module, used for distorting the first environment image to obtain a second fisheye image;

[0036] The space detection module is used to input the first fisheye image and the second fisheye image into a space detection model, and detect the accessible space of the vehicle's surrounding environment through the space detection model.

[0037] On the other hand, an embodiment of the present application provides a vehicle-mounted terminal, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor implements a method for detecting a reachable space as described in one aspect above and any optional implementation method thereof.

[0038] In another aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for detecting a reachable space as described in the above-mentioned other aspect and its optional manner is implemented.

[0039] The technical solution provided by the embodiments of the present application may at least include the following beneficial effects:

[0040] The vehicle-mounted terminal of the present application obtains a first fisheye image through a fisheye camera; obtains a first environment image through a shooting camera; distorts the first environment image to obtain a second fisheye image; inputs the first fisheye image and the second fisheye image into a space detection model, and detects the accessible space of the vehicle's surrounding environment through the space detection model. When using the space detection model to detect the accessible space, when the vehicle-mounted terminal captures the first fisheye image through its own fisheye camera, it also captures the first environment image through the shooting camera, and distorts the first environment image to obtain the second fisheye image, and inputs both the first fisheye image and the second fisheye image into the space detection model, and detects the accessible space of the vehicle's surrounding environment through the space detection model, so that the image data used by the vehicle-mounted terminal when performing the reachable space detection is more accurate, thereby improving the accuracy of the vehicle-mounted terminal's detection of the reachable space. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 is a schematic diagram of a fisheye image collected in a terminal according to an exemplary embodiment of the present application;

[0043] Figure 2 is a method flow chart of a method for detecting a reachable space provided by an exemplary embodiment of the present application;

[0044] Figure 3 is a schematic diagram of two images captured by a fisheye camera and a photographing camera according to an exemplary embodiment of the present application;

[0045] Figure 4 is a method flow chart of a method for detecting a reachable space involved in an exemplary embodiment of the present application;

[0046] Figure 5 is a schematic diagram of an image before and after distortion related to an exemplary embodiment of the present application;

[0047] Figure 6 An exemplary embodiment of the present application relates to Figure 5 An image schematic diagram of a target fisheye image;

[0048] Figure 7An exemplary embodiment of the present application relates to Figure 6 An image schematic diagram of a target fisheye image;

[0049] Figure 8 is a structural block diagram of a detection device for a reachable space provided by an exemplary embodiment of the present application;

[0050] Fig. 9 It is a structural schematic diagram of a vehicle-mounted terminal provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0051] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0052] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.

[0053] It should be noted that the terms "first", "second", "third" and "fourth" in the specification and claims of the present application are used to distinguish different objects rather than to describe a specific order. The terms "including" and "having" in the embodiments of the present application and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0054] The solution provided by the present application can be used in scenarios where accessible spaces are detected by using a terminal in daily life. To facilitate understanding, some special terms and application architectures involved in the embodiments of the present application are briefly introduced below.

[0055] Neural Networks (NN) is a complex network system formed by a large number of simple processing units (called neurons) that are widely interconnected. It reflects many basic characteristics of human brain functions and is a highly complex nonlinear dynamic learning system. Neural networks have large-scale parallel, distributed storage and processing, self-organization, self-adaptation and self-learning capabilities, and are particularly suitable for dealing with imprecise and fuzzy information processing problems that require considering many factors and conditions at the same time. The development of neural networks is related to neuroscience, mathematical science, cognitive science, computer science, artificial intelligence, information science, cybernetics, robotics, microelectronics, psychology, optical computing, molecular biology, etc., and is an emerging frontier interdisciplinary subject.

[0056] Global features refer to the overall properties of an image. Common global features include color features, texture features, and shape features, such as intensity histograms. Since they are low-level visual features at the pixel level, global features have good invariance, simple calculations, and intuitive representations. However, their high feature dimensions and large amount of calculations are their fatal weaknesses. In addition, global feature descriptions are not suitable for image aliasing and occlusion. Local features are features extracted from local areas of an image, including edges, corners, lines, curves, and areas with special attributes. Common local features include two types of descriptions: corner point type and area type.

[0057] Rectified Linear Unit (ReLU), also known as the rectified linear unit, is an activation function commonly used in artificial neural networks, usually referring to nonlinear functions represented by ramp functions and their variants.

[0058] In daily life, various terminals have been applied in various fields. For example, with the development of the vehicle field, vehicles are gradually moving towards autonomous driving. In the field of autonomous driving, the use of vehicle terminals is indispensable. Vehicles will encounter various obstacles, such as other vehicles, pedestrians, telephone poles, high-rise buildings and other obstacles. It is particularly important for cars to perceive the surrounding environment through vehicle terminals. In terms of environmental perception, data can be collected and calculated through hardware devices such as camera equipment and ultrasonic radar to achieve the effect of obstacle detection. For example, common camera equipment can include binocular cameras, monocular cameras, fisheye cameras and other devices.

[0059] Please refer to Figure 1 , which shows a schematic diagram of a fisheye image collected in a terminal according to an exemplary embodiment of the present application. Figure 1 As shown, the fisheye image 100 includes a first obstacle 101 and a second obstacle 102 .

[0060] Among them, the terminal in the present application is a terminal device that can be installed with the above-mentioned camera device, for example, a terminal that can be installed with a fisheye camera. The terminal can be a smart phone, a tablet computer, an e-book reader, smart glasses, a smart watch, an MP3 player (Moving Picture Experts Group Audio Layer III, Moving Picture Experts Compression Standard Audio Layer 3), an MP4 (Moving Picture Experts Group Audio Layer IV, Moving Picture Experts Compression Standard Audio Layer 4) player, a notebook computer, a laptop computer, a portable computer and a desktop computer, etc.

[0061] Optionally, the terminal is an IoT device that can be installed with the above-mentioned camera device, such as a sensor device, a computer of an IoT terminal, for example, it can be a fixed, portable, pocket-sized, handheld, computer-built-in or vehicle-mounted device. For example, station (STA), subscriber unit, subscriber station, mobile station, mobile, remote station, access point, remote terminal, access terminal, user terminal, user agent, user device. Alternatively, the terminal can also be a device of an unmanned aerial vehicle. Alternatively, the terminal can also be a vehicle-mounted device, for example, it can be a driving computer with an audio playback function, or a wireless communication device external to the driving computer.

[0062] In the vehicle field, technologies such as autonomous driving and space detection have been widely used. When performing space detection on a vehicle, developers are often required to pre-train a space detection model, pre-set the space detection model in the vehicle, and detect the accessible space through the space detection model. Optionally, during the model training process, developers can use a large number of image samples to train the model, for example, using 1,000 fisheye images for training. The trained space detection model is used to input a fisheye image, and the space detection model can output the corresponding space detection result.

[0063] Optionally, when the above model is applied to an actual vehicle, the fisheye image is usually collected by the vehicle's fisheye camera, and the accessible space of the vehicle's surrounding environment is detected based on the fisheye image. Usually, one fisheye image cannot fully reflect the accessible space of the vehicle's surrounding environment, and the fisheye camera needs to be photographed at different positions to complete the detection of the accessible space through the space detection model. For example, in the above Figure 1 In the fisheye image 100, the terminal can trigger the function of detecting the reachable space, wherein the reachable space generally refers to the space that can be reached in the image. For example, the vehicle-mounted terminal obtains the spatial range that it can move through the fisheye image. Usually, the terminal can not only detect the positions of the first obstacle 101 and the second obstacle 102 in the fisheye image, but also detect the properties of the first obstacle 101 and the second obstacle 102. Optionally, after obtaining the position coordinates of each obstacle in the reachable space, the terminal can also detect the properties of each obstacle in the reachable space in combination with the heat map (for example, the property categories of each obstacle are divided into several property categories such as static obstacles, dynamic obstacles, and suspended obstacles).

[0064] In the above-mentioned method of detecting the accessible space, in order to detect a more complete accessible space around the vehicle, the fisheye camera needs to continuously collect in different directions. When the vehicle-mounted terminal allows the fisheye camera to continuously collect in different directions, due to the movement of the vehicle, there will be a time difference between the fisheye images collected in other directions and the fisheye images collected for the first time. The collected fisheye images cannot be used in time, which is not conducive to real-time acquisition of the accessible space during the movement of the vehicle. If the accessible space of the vehicle's surrounding environment is detected by only obtaining a fisheye image once with the fisheye camera, the accuracy of the accessible space detection will be low due to the small number of features in the collected fisheye images.

[0065] In order to increase the image data used in reachable space detection and improve the accuracy of reachable space detection, the present application proposes a solution. By distorting the images taken by other cameras on the vehicle-mounted terminal, the corresponding fisheye images can also be obtained, and multiple fisheye images are simultaneously input into the space detection model to realize the detection of reachable space.

[0066] Please refer to Figure 2 , which shows a method flow chart of a method for detecting a reachable space provided by an exemplary embodiment of the present application. The reachable space detection method can be applied to the above Figure 1 The vehicle terminal in the scenario architecture shown in FIG. Figure 2 As shown, the method for detecting the reachable space may include the following steps.

[0067] Step 201: Acquire a first fisheye image through a fisheye camera.

[0068] Optionally, in the present application, the vehicle is equipped with a fisheye camera and a shooting camera, the fisheye camera can obtain a fisheye image by shooting, and the shooting camera can obtain a normal image by shooting. When performing space detection, the vehicle-mounted terminal can obtain the first fisheye image by controlling the fisheye camera.

[0069] Step 202: Acquire a first environment image by using a shooting camera.

[0070] Similarly, the vehicle terminal can also control the camera to obtain the first environment image. Figure 3 , which shows a schematic diagram of two images captured by a fisheye camera and a shooting camera involved in an exemplary embodiment of the present application. Figure 3 As shown, it includes Figure 3 (a) and Figure 3 (b) Figure 3 (a) includes a first fisheye image 301, Figure 3 (b) includes a first environment image 302 .

[0071] Step 203: distort the first environment image to obtain a second fisheye image.

[0072] Optionally, the vehicle-mounted terminal may distort the first environment image to obtain a second fisheye image. For example, the vehicle-mounted terminal may input the first environment image into a distortion model, and distort the first environment image obtained by the camera using a preset distortion model to obtain a second fisheye image. The distortion model may be obtained by the developer through pre-training using a normal image and a distorted fisheye image.

[0073] Step 204: input the first fisheye image and the second fisheye image into a space detection model, and detect the accessible space of the vehicle's surrounding environment through the space detection model.

[0074] The vehicle-mounted terminal uses the first fisheye image and the second fisheye image as inputs to the space detection model to detect the accessible space of the vehicle's surrounding environment. Through the space detection model, the vehicle-mounted terminal can obtain the accessible space of the vehicle's surrounding environment based on the first fisheye image and the second fisheye image, thereby achieving more accurate detection of the vehicle's accessible space.

[0075] In summary, the vehicle-mounted terminal of the present application obtains a first fisheye image through a fisheye camera; obtains a first environment image through a shooting camera; distorts the first environment image to obtain a second fisheye image; inputs the first fisheye image and the second fisheye image into a space detection model, and detects the accessible space of the vehicle's surroundings through the space detection model. When using the space detection model to detect the accessible space, when the vehicle-mounted terminal captures the first fisheye image through its own fisheye camera, it also captures the first environment image through the shooting camera, and distorts the first environment image to obtain the second fisheye image, and inputs both the first fisheye image and the second fisheye image into the space detection model to detect the accessible space of the vehicle's surroundings through the space detection model, so that the image data used by the vehicle-mounted terminal when performing the reachable space detection is more accurate, thereby improving the accuracy of the vehicle-mounted terminal's detection of the reachable space.

[0076] In one possible implementation, the space detection model in the vehicle-mounted terminal can perform feature fusion on the first fisheye image and the second fisheye image to obtain a target fusion image, and detect the accessible space of the vehicle's surrounding environment through the target fusion image.

[0077] Please refer to Figure 4 , which shows a method flow chart of a method for detecting a reachable space involved in an exemplary embodiment of the present application. The reachable space detection method can be applied to the above Figure 1 The vehicle terminal in the scenario architecture shown in FIG. Figure 4 As shown, the method for detecting the reachable space may include the following steps.

[0078] Step 401: Acquire a first fisheye image through a fisheye camera.

[0079] Step 402: Acquire a first environment image by using a shooting camera.

[0080] Optionally, the acquisition of the first fisheye image and the first environment image in the vehicle-mounted terminal is similar to that in the above-mentioned step 201 and step 202, and will not be repeated here.

[0081] Step 403: distort the first environment image to obtain a second fisheye image.

[0082] Optionally, the vehicle terminal obtains a first distorted image based on the first environment image and the first distortion function, performs coordinate screening on the first distorted image, determines a target area image in the first distorted image, and obtains a second fisheye image based on the target area image.

[0083] Optionally, a first distortion function may be pre-set in the vehicle terminal, and the vehicle terminal uses the first distortion function to distort the first environment image acquired by the shooting camera to obtain a first distorted image. For example, the first distortion function may be a function used in Opencv2 to distort an image: UndistortImage() function, initUndistortRectifyMap(), remap() function, etc. After the vehicle terminal distorts the first environment image, a first distorted image may be obtained, and the coordinates of each position in the first environment image are also changed due to the distortion.

[0084] For example, please refer to Figure 5 , which shows a schematic diagram of an image before and after distortion related to an exemplary embodiment of the present application. Figure 5 As shown in (a), the first environment image 501 is captured by the vehicle-mounted terminal controlling the camera. Figure 5 (b) shows a first distorted image 502 obtained by distorting the first environment image 501 by the vehicle-mounted terminal according to the first distortion function.

[0085] Optionally, in the present application, the vehicle terminal continues to perform coordinate screening on the first distorted image, determines the target area image in the first distorted image, and obtains the second fisheye image according to the determined target area image. In the first environment image, each pixel point is represented by a pixel coordinate (horizontal coordinate X and vertical coordinate Y). Since the original first environment image is distorted in both the X direction and the Y direction in the distorted first distorted image, each pixel point of the first environment image is compressed, expanded, etc., so that some of the pixels in the X direction overlap. For example, two pixels P1 (X1, Y1) and P2 (X2, Y2) in the first environment image correspond to P3 (X3, Y3) and P4 (X3, Y4) in the first distorted image after distortion. At this time, the vehicle terminal screens the coordinates according to the size of the Y coordinate, and removes a pixel point with a smaller Y coordinate value. For example, if Y3 is greater than Y4, the vehicle terminal can remove P4, thereby retaining P3.

[0086] Optionally, after the above-mentioned coordinate screening, the vehicle-mounted terminal can obtain the target area image in the first distorted image, and obtain the second fisheye image based on the determined target area image. In one possible implementation, the vehicle-mounted terminal enlarges and trims the target area image to obtain the target fisheye image; obtains a first added image based on the first fisheye image, the first added image being the blind area image in the first fisheye image; and adds the first added image to the target fisheye image to generate the second fisheye image.

[0087] Optionally, the vehicle terminal can enlarge the target area image through the resize function, and crop the enlarged target area image through the crop function to obtain the target fisheye image. Figure 6 , which shows that an exemplary embodiment of the present application involves Figure 5 A schematic diagram of a target fisheye image. Figure 6 As shown, in Figure 5 Image 601 is obtained by enlarging the target area image in (b) and trimming the surrounding black area.

[0088] Optionally, the vehicle terminal may also obtain a first additional image based on the first fisheye image, and add the first additional image to the target fisheye image to generate a second fisheye image. The first additional image is a blind area image in the first fisheye image. For example, the vehicle terminal may perform a cutout process on the first fisheye image in advance, extract the blind area image in the first fisheye image, and use the blind area image as the first additional image. After obtaining the target fisheye image, the vehicle terminal adds the first additional image to the target fisheye image in the form of splicing, overlaying, etc., to generate a second fisheye image. Please refer to Figure 7 , which shows that an exemplary embodiment of the present application involves Figure 6 A schematic diagram of a target fisheye image. Figure 7 As shown, it includes a second fisheye image 701, a first added image 702 and a target fisheye image 703. Figure 7 Yes Figure 6 A second fisheye image 701 is obtained by adding the first added image to the target fisheye image in FIG.

[0089] It should be noted that the above process in the present application can also be applied in the model training process. For example, the vehicle terminal can also provide a model training function (the model training function is used to train the above-mentioned space detection model). When the user triggers the model training function, the vehicle terminal needs to obtain sample fisheye images and train the model based on the sample fisheye images. Optionally, when the vehicle terminal determines that the number of sample fisheye images obtained is less than a preset threshold, the vehicle terminal can also process the previously stored environmental images or the environmental images taken by the camera through the above steps 402 to 403 to obtain fisheye images as sample images for model training, thereby increasing the amount of data for model training and improving the accuracy of model training. Among them, the preset threshold is pre-set in the vehicle terminal by the developer.

[0090] For example, the preset threshold is 500. When the user triggers the model training function, the number of sample fisheye images obtained by the vehicle terminal is 400. At this time, the number of sample fisheye images obtained by the vehicle terminal is less than the preset threshold. The vehicle terminal obtains the environmental image or the environmental image captured by the camera, and obtains the fisheye image through the above steps 402 and 403 of the environmental image, and uses the data in the sample fisheye image for training.

[0091] It should be noted that, in the above-mentioned model training process, for the sample fisheye images obtained based on the environmental image, in the process of obtaining the target area image, the training weight outside the target area image can be set to 0, and the training weight in the blind area image can be set to 0.1, thereby improving the training efficiency.

[0092] Step 404: input the first fisheye image and the second fisheye image into the space detection model, perform feature fusion on the first fisheye image and the second fisheye image through the space detection model, and obtain a target fused image.

[0093] In one possible implementation, the vehicle-mounted terminal can perform image recognition on the first fisheye image through the spatial detection model to obtain the first image feature; perform image recognition on the second fisheye image through the spatial detection model to obtain the second image feature; and fuse the first fisheye image and the second fisheye image according to the first image feature and the second image feature to obtain the target fused image. For example, in the spatial detection model, the image recognition module performs image recognition on the first fisheye image and the second fisheye image to obtain the first image feature corresponding to the first fisheye image and the second image feature corresponding to the second fisheye image. The vehicle-mounted terminal fuses the first fisheye image and the second fisheye image based on the first image feature and the second image feature to obtain the target fused image.

[0094] For example, after the spatial detection model performs image recognition on the first fisheye image, the first image features obtained include the position features of the first obstacle, the position features of the second obstacle, and the position features of the third obstacle. After the spatial detection model performs image recognition on the second fisheye image, the second image features obtained include the position features of the first obstacle, the position features of the third obstacle, and the position features of the fourth obstacle. In the world coordinate system, for the first obstacle and the third obstacle in the first fisheye image, and the first obstacle and the third obstacle in the second fisheye image, the first obstacle and the third obstacle are the same. Therefore, the image in the image coordinate system of the second fisheye image can be transformed into the image coordinate system of the first fisheye image, thereby obtaining a target fused image in which the second fisheye image is fused into the first fisheye image. Optionally, fusing the second fisheye image into the first fisheye image is also similar to this, which will not be repeated here.

[0095] Step 405 , based on the target fusion image, detect the accessible space of the vehicle surrounding environment through a space detection model.

[0096] In one possible implementation, the vehicle-mounted terminal can perform feature extraction on the target fusion image to obtain a feature pyramid, which includes image features of multiple layers of target fusion images; based on the image features of the target layer in the feature pyramid and a regression model, the initial boundary of the reachable space in the target fusion image is obtained, the target layer is the layer with the largest number of features in the feature pyramid, and the regression model is a model that is trained based on at least two sample layers and obtains an initial boundary with the same width as the sample layer; based on the initial boundary, the reachable space of the vehicle's surrounding environment is detected through a space detection model.

[0097] In the present application, the vehicle terminal uses the target residual network to obtain a multi-layer feature image based on the acquired target fusion image. The target residual network is a multi-layer feature pre-designed by the developer for extracting the target fusion image. For example, the target residual network can be a Resnet34 residual network. Taking the image parameters of the first fisheye image as (h1, w1, 3) as an example, after the vehicle terminal processes the target fusion image through the Resnet34 residual network, the multi-layer features obtained are as follows: (h1 / 4, w1 / 4, 64), (h1 / 8, w1 / 8, 128), (h1 / 16, w1 / 16, 256), (h1 / 32, w1 / 32, 512).

[0098] After the vehicle terminal obtains the multi-layer feature image, it generates a feature pyramid based on the multi-layer feature image and the pyramid network. The feature pyramid includes the image features of the multi-layer target fusion image. Optionally, the pyramid network is the same as the target residual network, and is also pre-designed by the developer and set in the vehicle terminal. For example, the pyramid network can be a feature pyramid network (FPN), in which FPN is a top-down feature fusion method, and FPN is a multi-scale target detection algorithm, that is, there is more than one feature prediction layer. Although some algorithms also use multi-scale feature fusion for target detection, they often only use the features of one scale obtained after fusion. Although this approach can fuse the semantic information of the top-level features and the detail information of the bottom-level features, it will cause some deviations in the process of feature deconvolution, etc., and only using the features obtained after fusion for prediction will have an adverse effect on the detection accuracy. Starting from the above problem, the FPN method can make predictions on fused features of multiple different scales to maximize the detection accuracy.

[0099] Optionally, after obtaining the feature pyramid, the vehicle terminal extracts a layer containing the largest number of features in the feature pyramid (i.e., the target layer), inputs the target layer into the regression model, and uses the regression model to obtain the initial boundary of the reachable space in the target fused image.

[0100] In one possible implementation, the vehicle terminal extracts the boundary of the target layer according to the regression model to obtain the boundary of the reachable space in the target layer; the boundary of the reachable space in the target layer is used as the initial boundary of the reachable space in the target fusion image, and the initial boundary is a single-channel array, and the width of the initial boundary is the same as the width of the target layer. For example, the vehicle terminal uses the first convolution layer and the second convolution layer in the regression model to process the target layer in sequence to extract the initial boundary of the reachable space in the target layer.

[0101] Optionally, taking the above-mentioned target fusion image image parameters as (h1, w1, 3), after obtaining the corresponding feature pyramid, the image parameters of the target layer are (h1 / 4, w1 / 4, 128) as an example, the first convolutional layer can be: conv(128*64*3*3), where each number in conv represents: output channel, input channel, channel width (kernel width), channel height (kernel height). The vehicle terminal convolves the target layer with the first convolutional layer, and performs batch normalization (BN) and the first ReLU function on the convolution result to obtain a regression feature map, where the image parameters of the regression feature map are the same as those of the target layer. The image parameters of the regression feature map are also (h1 / 4, w1 / 4, 128).

[0102] Optionally, after obtaining the regression feature map in the regression model, the regression feature map is processed by the second convolution layer to extract the initial boundary of the reachable space in the target layer. Among them, the second convolution layer can be: conv(320*128*w1 / 4*h1 / 4), the regression model convolves the regression feature map obtained above with the second convolution layer, and performs BN and the second ReLU function processing on the convolution result in turn to obtain the initial boundary of the reachable space in the target layer, and the obtained initial boundary of the reachable space in the target layer is used as the initial boundary of the reachable space in the target fusion image. Optionally, the initial boundary can be represented by y_init, then, based on the above example, y_init obtained in this step is a single-channel array, and the width of the initial boundary is the same as the width of the target layer, both of which are w1 / 4. The vehicle-mounted terminal determines the initial boundary as the actual boundary of the reachable space in the vehicle's surrounding environment.

[0103] In one possible implementation, after detecting the reachable space of the vehicle's surrounding environment through the space detection model based on the target fusion image, the vehicle-mounted terminal can also obtain an attribute feature image based on the target fusion image, where the attribute feature image includes attribute features of each obstacle in the target fusion image; classify the attribute feature image to obtain an attribute map corresponding to each obstacle attribute, where the attribute map includes the confidence of the obstacle attribute corresponding to each pixel point in the target fusion image; and determine the obstacle attribute of the obstacle at a first position in the target fusion image based on the attribute map, where the first position is a pixel point position contained in any obstacle in the target fusion image.

[0104] For example, the vehicle terminal can determine the shape features, position features, etc. of each obstacle in the target fusion image based on the target fusion image, and obtain the corresponding attribute features according to the shape features and position features of each obstacle, thereby generating an attribute feature image including each obstacle. For example, the image parameters of the target fusion image are (h1, w1, 3), where h1 represents the height of the target fusion image, w1 represents the width of the target fusion image, and 3 represents that each pixel is 3 channels. After extracting the attribute features of the target fusion image, the attribute feature image can be as follows: (h1 / 4, w1 / 4, 128).

[0105] Optionally, the obstacle attribute categories of each obstacle are divided into several attribute categories, namely, static obstacles, dynamic obstacles, and suspended obstacles. The vehicle-mounted terminal can obtain the attribute map corresponding to the static obstacles, and the attribute map corresponding to the static obstacles includes the obstacle attribute of each obstacle in the target fusion image is a static obstacle, and the confidence corresponding to the obstacle attribute of each obstacle is a static obstacle. The vehicle-mounted terminal can also obtain the attribute map corresponding to the dynamic obstacles, and the attribute map corresponding to the dynamic obstacles includes the obstacle attribute of each obstacle in the target fusion image is a dynamic obstacle, and the confidence corresponding to the obstacle attribute of each obstacle is a dynamic obstacle. The vehicle-mounted terminal can also obtain the attribute map corresponding to the suspended obstacles, and the attribute map corresponding to the suspended obstacles includes the obstacle attribute of each obstacle in the target fusion image is a suspended obstacle, and the confidence corresponding to the obstacle attribute of each obstacle is a dynamic obstacle.

[0106] Optionally, the vehicle terminal can also obtain an attribute vector based on the attribute feature image, and the attribute vector is based on the obstacle attributes and confidence corresponding to each pixel in the target fusion image under global classification. Among them, global classification can refer to the classification result obtained by the vehicle terminal after inputting the attribute feature image into the global feature obtained by the fully connected layer. The attribute vector includes not only the classification result but also the confidence of the classification result. For example, the obstacle attribute of an obstacle at one position is a suspended obstacle, and the attribute vector also includes the confidence that the obstacle at this position is a suspended obstacle.

[0107] Optionally, the vehicle terminal determines the obstacle attribute corresponding to the pixel position of any obstacle in the target fused image according to the attribute vector obtained above and the attribute map corresponding to each obstacle attribute. That is, in the method of classifying the attribute feature image and obtaining the attribute map corresponding to each obstacle attribute, the vehicle terminal obtains the obstacle attribute corresponding to each obstacle and the confidence of the obstacle attribute. The vehicle terminal can also obtain the attribute vector representing the obstacle attribute and the confidence corresponding to each pixel in the target fused image under the global classification according to the attribute feature image based on the fully connected layer, and combine the two classification results to determine the obstacle attribute of the obstacle at the first position in the target fused image.

[0108] Optionally, the first position may be a boundary position of an obstacle determined by the vehicle terminal through marking or calculation. For example, the vehicle terminal determines the boundary of any obstacle through a boundary determination model, and uses any position on the boundary as the first position. The boundary determination model may be pre-trained by the developer using different fisheye images and set in the vehicle terminal.

[0109] In summary, the vehicle-mounted terminal of the present application obtains a first fisheye image through a fisheye camera; obtains a first environment image through a shooting camera; distorts the first environment image to obtain a second fisheye image; inputs the first fisheye image and the second fisheye image into a space detection model, and detects the accessible space of the vehicle's surroundings through the space detection model. When using the space detection model to detect the accessible space, when the vehicle-mounted terminal captures the first fisheye image through its own fisheye camera, it also captures the first environment image through the shooting camera, and distorts the first environment image to obtain the second fisheye image, and inputs both the first fisheye image and the second fisheye image into the space detection model to detect the accessible space of the vehicle's surroundings through the space detection model, so that the image data used by the vehicle-mounted terminal when performing the reachable space detection is more accurate, thereby improving the accuracy of the vehicle-mounted terminal's detection of the reachable space.

[0110] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0111] Please refer to Figure 8 , which shows a structural block diagram of a detection device for a reachable space provided by an exemplary embodiment of the present application. The detection device for a reachable space 800 can be applied to a vehicle-mounted terminal. The detection device for a reachable space includes:

[0112] A first acquisition module 801 is used to acquire a first fisheye image through a fisheye camera;

[0113] The second acquisition module 802 is used to acquire a first environment image through a shooting camera;

[0114] A third acquisition module 803 is used to distort the first environment image to obtain a second fisheye image;

[0115] The space detection module 804 is used to input the first fisheye image and the second fisheye image into a space detection model, and detect the accessible space of the vehicle's surrounding environment through the space detection model.

[0116] In summary, the vehicle-mounted terminal of the present application obtains a first fisheye image through a fisheye camera; obtains a first environment image through a shooting camera; distorts the first environment image to obtain a second fisheye image; inputs the first fisheye image and the second fisheye image into a space detection model, and detects the accessible space of the vehicle's surroundings through the space detection model. When using the space detection model to detect the accessible space, when the vehicle-mounted terminal captures the first fisheye image through its own fisheye camera, it also captures the first environment image through the shooting camera, and distorts the first environment image to obtain the second fisheye image, and inputs both the first fisheye image and the second fisheye image into the space detection model to detect the accessible space of the vehicle's surroundings through the space detection model, so that the image data used by the vehicle-mounted terminal when performing the reachable space detection is more accurate, thereby improving the accuracy of the vehicle-mounted terminal's detection of the reachable space.

[0117] Optionally, the space detection module 804 includes: a feature fusion unit and a space detection unit;

[0118] The feature fusion unit is used to input the first fisheye image and the second fisheye image into a space detection model, perform feature fusion on the first fisheye image and the second fisheye image through the space detection model, and obtain a target fused image;

[0119] The space detection unit is used to detect the accessible space of the vehicle's surrounding environment based on the target fusion image.

[0120] Optionally, the feature fusion unit is used to

[0121] Performing image recognition on the first fisheye image by using the spatial detection model to obtain a first image feature;

[0122] Performing image recognition on the second fisheye image by using the spatial detection model to obtain second image features;

[0123] The first fisheye image and the second fisheye image are fused according to the first image feature and the second image feature to obtain the target fused image.

[0124] Optionally, the space detection unit is used to

[0125] Extracting features from the target fused image to obtain a feature pyramid, wherein the feature pyramid includes multiple layers of image features of the target fused image;

[0126] According to the image features of the target layer in the feature pyramid and the regression model, an initial boundary of the reachable space in the target fused image is obtained, wherein the target layer is a layer in the feature pyramid containing the largest number of features, and the regression model is a model trained based on at least two sample layers and obtaining an initial boundary with the same width as the sample layer;

[0127] According to the initial boundary, a reachable space of the vehicle surrounding environment is detected.

[0128] Optionally, the device further comprises:

[0129] A fourth acquisition module, configured to acquire an attribute feature image according to the target fused image after detecting the reachable space of the vehicle surrounding environment according to the target fused image, wherein the attribute feature image includes attribute features of each obstacle in the target fused image;

[0130] A fifth acquisition module, configured to classify the attribute feature image and obtain an attribute map corresponding to each obstacle attribute, wherein the attribute map includes a confidence level of the obstacle attribute corresponding to each pixel point in the target fused image;

[0131] The attribute determination module is used to determine the obstacle attribute of the obstacle at a first position in the target fused image according to the attribute map, where the first position is a pixel point position included in any obstacle in the target fused image.

[0132] Optionally, the third acquisition module includes: a first acquisition unit, a region determination unit and a second acquisition unit;

[0133] The first acquisition unit is used to acquire a first distorted image according to the first environment image and a first distortion function;

[0134] The region determination unit is used to perform coordinate screening on the first distorted image to determine a target region image in the first distorted image;

[0135] The second acquisition unit is used to acquire the second fisheye image according to the target area image.

[0136] Optionally, the second acquisition unit is used to

[0137] Performing magnification and edge trimming processing on the target area image to obtain a target fisheye image;

[0138] Acquire a first added image according to the first fisheye image, where the first added image is a blind area image in the first fisheye image;

[0139] The first added image is added to the target fisheye image to generate the second fisheye image.

[0140] Fig. 9 Schematic diagram of a vehicle terminal provided by an exemplary embodiment of the present application. Fig. 9 As shown, the vehicle-mounted terminal 900 includes a central processing unit (CPU) 901, a system memory 904 including a random access memory (RAM) 902 and a read-only memory (ROM) 903, and a system bus 905 connecting the system memory 904 and the central processing unit 901. The vehicle-mounted terminal 900 also includes a basic input / output system (I / O system) 908 for helping to transmit information between various components in the computer, and a large-capacity storage device 907 for storing an operating system 912, application programs 913 and other program modules 914.

[0141] The basic input / output system 906 includes a display 908 for displaying information and an input device 909 such as a mouse and a keyboard for user inputting information. The display 908 and the input device 909 are connected to the central processing unit 901 through an input / output controller 910 connected to the system bus 905. The basic input / output system 906 may also include an input / output controller 910 for receiving and processing inputs from a plurality of other devices such as a keyboard, a mouse, or an electronic stylus. Similarly, the input / output controller 910 also provides output to a display screen, a printer, or other types of output devices.

[0142] The mass storage device 907 is connected to the central processing unit 901 via a mass storage controller (not shown) connected to the system bus 905. The mass storage device 907 and its associated computer-readable medium provide non-volatile storage for the vehicle-mounted terminal 900. That is, the mass storage device 907 may include a computer-readable medium (not shown) such as a hard disk or a CD-ROM (Compact Disc Read-Only Memory) drive.

[0143] The computer-readable medium may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Computer storage media include RAM, ROM, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory or other solid-state storage technologies, CD-ROM, DVD (Digital Video Disc) or other optical storage, cassettes, tapes, disk storage or other magnetic storage devices. Of course, those skilled in the art will know that the computer storage media is not limited to the above. The above-mentioned system memory 904 and mass storage device 907 can be collectively referred to as memory.

[0144] The vehicle-mounted terminal 900 can be connected to the Internet or other network devices through a network interface unit 911 connected to the system bus 905 .

[0145] The memory also includes one or more programs, which are stored in the memory. The central processing unit 901 implements all or part of the steps performed by the vehicle terminal in the methods provided in the above embodiments of the present application by executing the one or more programs.

[0146] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0147] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, a computer, a server, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or a data center that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a high-density digital video disc (Digital Video Disc, DVD)), or a semiconductor medium (e.g., a solid-state drive (Solid State Disk, SSD)), etc.

[0148] The embodiment of the present application also discloses a vehicle, which includes an on-board terminal, the on-board terminal includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor implements the detection method of the reachable space in the above method embodiment. Optionally, the above terminal can be the on-board terminal in this embodiment.

[0149] The embodiment of the present application further discloses a computer-readable storage medium storing a computer program, wherein the computer program implements the method in the above method embodiment when executed by a processor.

[0150] It should be understood that the "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. Those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required for the present application.

[0151] In the various embodiments of the present application, it should be understood that the size of the serial numbers of the above-mentioned processes does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0152] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0153] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0154] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-accessible memory. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a memory and includes several requests for a computer device (which can be a personal computer, a server or a network device, etc., specifically a processor in a computer device) to perform some or all of the steps of the above-mentioned methods of various embodiments of the present application.

[0155] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0156] The above is an example introduction to a detection method, device, vehicle-mounted terminal and storage medium for a reachable space disclosed in an embodiment of the present application. Individual examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for detecting a reachable space, characterized in that: The method is applied to a vehicle-mounted terminal, and the method comprises: Acquire a first fisheye image by using a fisheye camera; Acquire a first environment image by shooting a camera; The first environment image is distorted to obtain a second fisheye image; wherein the second fisheye image is generated by adding a first added image to a target fisheye image, the first added image is an image of a blind area in the first fisheye image, and the target fisheye image is determined after coordinate screening of a first distorted image obtained by distorting the first environment image; The first fisheye image and the second fisheye image are input into a space detection model, and the accessible space of the vehicle surrounding environment is detected by the space detection model.

2. The method according to claim 1, characterized in that The step of inputting the first fisheye image and the second fisheye image into a space detection model and detecting the accessible space of the surrounding environment of the vehicle by using the space detection model includes: Inputting the first fisheye image and the second fisheye image into a spatial detection model, performing feature fusion on the first fisheye image and the second fisheye image through the spatial detection model, and acquiring a target fused image; According to the target fused image, a reachable space of the vehicle surrounding environment is detected.

3. The method according to claim 2, characterized in that The step of inputting the first fisheye image and the second fisheye image into a space detection model, and performing feature fusion on the first fisheye image and the second fisheye image through the space detection model to obtain a target fused image includes: Performing image recognition on the first fisheye image by using the spatial detection model to obtain a first image feature; Performing image recognition on the second fisheye image by using the spatial detection model to obtain second image features; The first fisheye image and the second fisheye image are fused according to the first image feature and the second image feature to obtain the target fused image.

4. The method according to claim 2, characterized in that: The detecting the accessible space of the vehicle surrounding environment according to the target fusion image includes: Extracting features from the target fused image to obtain a feature pyramid, wherein the feature pyramid includes multiple layers of image features of the target fused image; According to the image features of the target layer in the feature pyramid and the regression model, an initial boundary of the reachable space in the target fused image is obtained, wherein the target layer is a layer in the feature pyramid containing the largest number of features, and the regression model is a model trained based on at least two sample layers and obtaining an initial boundary with the same width as the sample layer; According to the initial boundary, a reachable space of the vehicle surrounding environment is detected.

5. The method according to any one of claims 2 to 4, characterized in that: After detecting the reachable space of the vehicle surrounding environment according to the target fusion image, the method further includes: According to the target fused image, an attribute feature image is acquired, wherein the attribute feature image includes attribute features of each obstacle in the target fused image; Performing attribute classification on the attribute feature image to obtain an attribute graph corresponding to each obstacle attribute, wherein the attribute graph includes the confidence of the obstacle attribute corresponding to each pixel point in the target fused image; According to the attribute map, an obstacle attribute of an obstacle at a first position in the target fused image is determined, where the first position is a pixel point position included in any obstacle in the target fused image.

6. The method according to claim 1, characterized in that The step of distorting the first environment image to obtain a second fisheye image includes: Acquire a first distorted image according to the first environment image and a first distortion function; Performing coordinate screening on the first distorted image to determine a target area image in the first distorted image; The second fisheye image is acquired according to the target area image.

7. The method according to claim 6, characterized in that The step of acquiring the second fisheye image according to the target area image includes: Performing magnification and edge trimming processing on the target area image to obtain a target fisheye image; Acquire a first added image according to the first fisheye image, where the first added image is a blind area image in the first fisheye image; The first added image is added to the target fisheye image to generate the second fisheye image.

8. A detection device for accessible space, characterized in that: The device is applied to a vehicle-mounted terminal, and the device comprises: A first acquisition module, used for acquiring a first fisheye image through a fisheye camera; A second acquisition module, used for acquiring a first environment image through a shooting camera; a third acquisition module, configured to distort the first environment image to obtain a second fisheye image; wherein the second fisheye image is generated by adding a first added image to a target fisheye image, the first added image is an image of a blind area in the first fisheye image, and the target fisheye image is determined after coordinate screening of a first distorted image obtained by distorting the first environment image; The space detection module is used to input the first fisheye image and the second fisheye image into a space detection model, and detect the accessible space of the vehicle's surrounding environment through the space detection model.

9. A vehicle-mounted terminal, characterized in that: The vehicle-mounted terminal includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor implements the method for detecting a reachable space as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for detecting a reachable space as described in any one of claims 1 to 7 is implemented.

Images