Image display mode switching method and vehicle

Abstract

This disclosure provides a display mode switching method and a vehicle. The image display mode switching method includes: in response to a trigger signal that switches a display device from a current display mode to a target display mode, executing a first processing task and a second processing task in parallel, wherein the first processing task initializes image processing resources dependent on the target display mode, and the second processing task activates image acquisition resources dependent on the target display mode, and the first and second processing tasks overlap at least partially in timing; after both tasks are completed, generating and outputting a display image corresponding to the target display mode based on image data acquired by the image acquisition resources and the processing results of the image processing resources. By executing the image processing resource initialization and image acquisition resource activation in parallel, the total image loading time during display mode switching is reduced, and the switching response speed is improved.

Description

Technical Field

[0001] This disclosure relates to the field of display technology, specifically to an image display mode switching method and a vehicle. Background Technology

[0002] In instrument cluster display systems, some systems support switching between multiple display modes. Among these, the thermal imaging mode can provide drivers with auxiliary visibility at night or in low-visibility conditions. However, in related technologies, when the instrument cluster switches from other display modes to thermal imaging mode, the image loading time is relatively long, often approaching one second. This means that users cannot immediately obtain the thermal image after the switch, affecting the user experience. Summary of the Invention

[0003] In view of this, embodiments of the present disclosure provide an image display mode switching method and a vehicle to solve the problem of excessive image loading delay when switching display modes.

[0004] In a first aspect, a method for switching image display modes is provided, comprising: in response to a trigger signal that switches a display device from a current display mode to a target display mode, executing a first processing task and a second processing task in parallel, wherein the first processing task initializes image processing resources on which the target display mode depends, and the second processing task enables image acquisition resources on which the target display mode depends, and the first processing task and the second processing task overlap at least partially in timing; after both the first processing task and the second processing task have been completed, generating a display image corresponding to the target display mode based on image data acquired by the image acquisition resources and processing results of the image processing resources and outputting it to the display device.

[0005] The image display mode switching method provided in the above embodiments sets the initialization of image processing resources and the activation of image acquisition resources as tasks that are at least partially executed in parallel. This allows the time-consuming operations that were originally superimposed in sequence to be performed in an overlapping manner, eliminating the extra waiting time between tasks. As a result, the total time from triggering to switching to displaying an image is significantly shortened, effectively improving the switching response speed and enhancing the user experience.

[0006] In some embodiments, performing a first processing task and a second processing task includes: performing the initialization of a basic function software package corresponding to a target display mode in parallel; and performing the first processing task and the second processing task in parallel after the initialization of the basic function software package is completed.

[0007] The image display mode switching method provided in the above embodiments initializes the basic function software package as a prerequisite step, providing a common basic operating environment for the first and second processing tasks, enabling the first and second processing tasks to execute in parallel. This eliminates the serial blocking of image acquisition resource activation during image processing resource initialization, allowing two independent tasks to be started simultaneously after the basic function software package initialization is complete, compressing the original total time of serial aggregation into a shorter time after parallel execution.

[0008] In some embodiments, performing the second processing task includes: issuing an enable command to the image acquisition resource and obtaining image data fed back by the image acquisition resource; after confirming that the image data meets the target availability conditions, if the first processing task has not yet been completed, keeping the image acquisition resource in the enabled state and waiting for the first processing task to be completed.

[0009] The image display mode switching method provided in the above embodiments can ensure that the image acquisition resources will not be shut down or restarted due to the incomplete first processing task after they are ready, but will remain on and wait for the first processing task to be completed. Thus, image generation and output can be performed immediately after both tasks are completed, avoiding additional restart delays.

[0010] In some embodiments, confirming that the image data meets the target availability conditions includes: receiving a normal flag bit reported by the image acquisition resource, wherein the normal flag bit is used to indicate that the image acquisition resource has completed the shutter action and the output image data quality is normal; and confirming that the image data meets the target availability conditions based on the normal flag bit.

[0011] The image display mode switching method provided in the above embodiments provides a clear judgment mechanism based on the status feedback of the image acquisition resource itself, enabling the system to accurately determine whether the image acquisition resource has been started and output image data of normal quality.

[0012] In some embodiments, the image display mode switching method further includes: preprocessing the image data fed back by the image acquisition resource before the first processing task has been completed; generating a display image corresponding to the target display mode based on the image data acquired by the image acquisition resource and the processing result of the image processing resource, including: generating a display image based on the preprocessed image data and the processing result of the image processing resource.

[0013] The image display mode switching method provided in the above embodiments can fully utilize idle computing resources while waiting for the first processing task to complete by shifting some image processing work forward. After the first processing task is completed, the preprocessed image data is input to the initialized image processing resource for processing, and the final display image is generated based on the processing result. Since the image data has been preprocessed, the image processing resource can directly perform core processing after receiving the preprocessed image data without performing basic image preprocessing steps, thereby shortening the time from receiving data to outputting processing results and making the generation of the final display image more efficient.

[0014] In some embodiments, the target display mode is night vision mode, the image processing resource is an artificial intelligence image recognition model, and the image acquisition resource is an infrared camera.

[0015] The image display mode switching method provided in the above embodiments executes the time-consuming AI model initialization and the activation of the infrared camera in parallel in night vision mode. Since the loading time of the AI ​​model is usually much longer than the time required for the camera to be activated and the image to be stabilized, the parallel execution of the two processes effectively reduces the image loading waiting time when the instrument switches to night vision mode.

[0016] In some embodiments, performing the first processing task includes: loading model parameters of the artificial intelligence image recognition model; and initializing the inference runtime environment of the artificial intelligence image recognition model on the vehicle chip.

[0017] The image display mode switching method provided in the above embodiments incorporates the initialization of the inference runtime environment and the loading of model parameters into the first processing task and executes them in parallel with the second processing task, thereby reducing the loading delay of the night vision mode.

[0018] In some embodiments, the start time of the first processing task is the same as the start time of the second processing task.

[0019] The image display mode switching method provided in the above embodiments has a total switching time equal to the time of the longer of the two tasks, rather than the sum or partial superposition of the two, thus further achieving time compression.

[0020] In some embodiments, the image display mode switching method is applied to a controller; before responding to a trigger signal that switches the display device from the current display mode to the target display mode, the method further includes: when the display device is in the current display mode, if the controller's processing resources meet a preset idle condition, preloading at least a portion of the initialization data of the image processing resources on which the target display mode depends; and performing a first processing task, including: completing the initialization of the image processing resources based on the preloaded initialization data.

[0021] The image display mode switching method provided in the above embodiments reduces the workload of the first processing task during actual execution and shortens the execution time through the preloading process of initialization data. In scenarios where the first processing task is executed in parallel with the second processing task, the shortening of the first processing task's execution time means that both parallel tasks are completed earlier, thereby further compressing the total time of display mode switching and making the user-perceived mode switching delay shorter.

[0022] In a second aspect, a computer-readable storage medium is provided, on which a computer program is stored, wherein when the computer program is executed by a processor, it implements the image display mode switching method provided in any of the first aspects above.

[0023] Thirdly, an image display mode switching device is provided, comprising: an execution module, configured to execute a first processing task and a second processing task in response to a trigger signal that switches a display device from a current display mode to a target display mode, wherein the first processing task is the initialization of image processing resources on which the target display mode depends, and the second processing task is the activation of image acquisition resources on which the target display mode depends, and the first processing task and the second processing task are executed in at least partially parallel order; and a display module, configured to generate a display image corresponding to the target display mode and output it to the display device based on the image data acquired by the image acquisition resources and the processing results of the image processing resources after both the first processing task and the second processing task have been completed.

[0024] Fourthly, a vehicle is provided, comprising: a processor; a memory for storing processor-executable instructions; and a processor for executing the display mode switching method provided in the first aspect. Attached Figure Description

[0025] Figure 1 The diagram shown is an architectural schematic of an image display system provided in an embodiment of this disclosure.

[0026] Figure 2 The diagram shown is a flowchart illustrating an image display mode switching method provided in an embodiment of this disclosure.

[0027] Figure 3 The diagram shown is a timing flow chart of an image display mode switching method in related technologies.

[0028] Figure 4 The diagram shown is a timing flowchart of an image display mode switching method provided in an embodiment of this disclosure.

[0029] Figure 5 The diagram shown is a structural schematic of an image display mode switching device provided in an embodiment of this disclosure.

[0030] Figure 6The diagram shown is a structural schematic of a vehicle provided in an embodiment of this disclosure. Detailed Implementation

[0031] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0032] In-vehicle instrument cluster systems typically integrate multiple display modes to meet the information needs of drivers in different driving scenarios. Common display modes include, but are not limited to: a navigation mode for displaying navigation maps and route guidance information; an entertainment mode for presenting multimedia playback information and vehicle status information; and a vehicle diagnostic and warning mode for displaying information such as tire pressure, braking system status, and engine fault codes. With the continuous development of driver assistance technologies, some vehicle models also integrate a night vision mode. This mode uses a night vision camera to capture thermal radiation images of the road ahead, processes them, and displays them on the instrument cluster screen. This effectively assists drivers in identifying pedestrians, animals, and obstacles at night or in low-visibility conditions such as rain, fog, or dust, significantly improving driving safety.

[0033] Drivers can freely switch between various display modes according to the actual driving environment and personal preferences. In related technologies, to achieve display mode switching, it is usually necessary to complete the preparatory tasks required for the corresponding display mode. For example, after receiving the night vision mode switching trigger signal, the system needs to first initialize the functional components. After initialization is complete, it sends an activation command to the night vision camera and waits for it to output stable and usable image data.

[0034] Under this switching mechanism, the time consumed by each task is cumulatively added, resulting in a corresponding increase in the total time from triggering to the final display of the image. In particular, when the initialization of functional components involves time-consuming operations such as loading complex models or building inference environments, the camera must wait for all preceding stages to be completed before it can be turned on, causing a significant delay in image loading when switching display modes, which affects the user experience.

[0035] For example, in the aforementioned night vision mode, to ensure the accuracy of target recognition, its functional components include a basic functional software package for night vision algorithms and a large-scale image recognition model. The initialization of the image recognition model requires completing processes such as device creation, main model loading, and small model loading, which takes a considerable amount of time. After the trigger signal is issued, it needs to go through the complete process of initializing the basic functional software package, initializing the image recognition model, and turning on the night vision camera. Moreover, the above steps are executed sequentially, and the total image loading time can be as long as one second.

[0036] To address the aforementioned technical problems, this disclosure provides an image display mode switching method, comprising: responding to a trigger signal that switches a display device from a current display mode to a target display mode, executing a first processing task and a second processing task in parallel, wherein the first processing task initializes the image processing resources on which the target display mode depends, and the second processing task activates the image acquisition resources on which the target display mode depends, the first processing task and the second processing task at least partially overlap in timing; after both the first processing task and the second processing task are completed, generating a display image corresponding to the target display mode based on the image data acquired by the image acquisition resources and the processing results of the image processing resources and outputting it to the display device, thereby shortening the total time from the trigger of the display mode switching to the output of the display image.

[0037] Figure 1 The diagram shown is an architectural schematic of an image display system provided in an embodiment of this disclosure.

[0038] like Figure 1 As shown, this implementation environment is applied in a vehicle with night vision assistance function, and mainly includes a display device 110, a control unit 120, and an image acquisition unit 130. The above units are interconnected through an in-vehicle communication harness to form a complete signal path for image acquisition, processing, and display.

[0039] The control unit 120 is the core computing component of this implementation environment, housing an in-vehicle infotainment chip and the in-vehicle infotainment software running on that chip. Optionally, the in-vehicle infotainment chip is an automotive-grade system-on-a-chip with neural network computing capabilities, providing an inference environment for image recognition algorithms. The control unit 120 is connected to the display device 110 via a display interface to output the image to be rendered to the display device 110; it is also connected to the image acquisition unit 130 via a wiring harness to send control commands to the image acquisition unit 130 and receive the image data it transmits.

[0040] Display device 110 is located in the driver's field of vision, such as an instrument display screen or a central control display screen, and is used to present images to the user in different display modes. The display modes include the conventional instrument display mode, as well as the driver assistance display mode generated based on images acquired by image acquisition unit 130. The user or vehicle control strategy can switch between multiple display modes according to the driving environment.

[0041] The image acquisition unit 130 is used to acquire images of the environment in front of or around the vehicle and transmit the acquired raw image data back to the control unit 120. Optionally, the image acquisition unit 130 includes a night vision camera, a visible light camera, etc. The image acquisition unit 130 and the control unit 120 can be connected via a low voltage differential signaling (LVDS) harness.

[0042] The control unit 120 runs several functional software modules, including a basic function software package and an image recognition model. The basic function software package is used to configure and manage the image acquisition unit 130, and to receive, encapsulate, and distribute the returned image data. The image recognition model is used to perform recognition, rendering, and other processing on the image data.

[0043] The control unit 120, based on the trigger signal for display mode switching generated by environmental information or user operation, schedules the basic function software package, image recognition model, and image acquisition unit 130 to perform corresponding initialization and data acquisition actions.

[0044] It is understood that the image display mode switching method provided in this disclosure is not only applicable to vehicle-mounted image display systems, but also to other scenarios requiring rapid display mode switching, including but not limited to security monitoring display systems and drone aerial photography display systems. In any scenario where image processing resource initialization and image acquisition resource activation are required during display mode switching, the technical solution of this disclosure can be adopted to shorten the overall time of mode switching and improve the user experience.

[0045] The above-described implementation environment provides the hardware and software foundation for the image display mode switching method described in the subsequent embodiments. The image display mode switching method provided by the embodiments of this disclosure is described in detail below with reference to the accompanying drawings.

[0046] Figure 2 The diagram shown is a flowchart illustrating an image display mode switching method according to an embodiment of this disclosure. Figure 2 As shown in the embodiments of this disclosure, the image display mode switching method includes the following steps.

[0047] S210, in response to a trigger signal that switches the display device from the current display mode to the target display mode, the first processing task and the second processing task are executed in parallel.

[0048] The trigger signal can be, for example, a mode switching command generated by the vehicle's software in response to user operation or changes in vehicle status. Alternatively, the trigger signal can be a signal that is automatically triggered based on preset conditions, such as a signal that automatically enters a specific display mode after the vehicle is powered on. The current display mode can be, for example, the conventional display mode of the vehicle's instrument panel, such as the instrument panel information display mode or the navigation display mode. The target display mode is the display mode to be presented after switching, and this target display mode depends on the collaborative work of specific image processing resources and image acquisition resources.

[0049] The first processing task is to initialize the image processing resources on which the target display mode depends, and the second processing task is to enable the image acquisition resources on which the target display mode depends.

[0050] Image processing resources can be software or hardware resources that require initialization before use, such as artificial intelligence (AI) image recognition models, image enhancement algorithm modules, or image fusion processing modules. Managing image processing resources as independent initialization tasks allows their initialization process to be scheduled in parallel with other tasks, preventing the entire switching process from being blocked due to the long initialization time of image processing resources.

[0051] Image acquisition resources can be, for example, infrared cameras, visible light cameras, or depth cameras—image acquisition devices that need to be activated before use. Activating the image acquisition resource is managed as an independent processing task, allowing its activation process to run in parallel with the initialization of image processing resources. This ensures that the preparation for activating the image acquisition resource is completed synchronously during the initialization of image processing resources.

[0052] The first and second processing tasks overlap at least partially in timing, meaning that the execution time periods of the first and second processing tasks overlap. For example, the first and second processing tasks can be started simultaneously at the same time, or the second processing task can be started after the first processing task has started but before it has been completed, so that their execution processes partially overlap in time.

[0053] This parallel execution strategy compresses the time consumed by two tasks that were originally executed serially, thereby effectively shortening the overall waiting time for mode switching.

[0054] S220: After both the first and second processing tasks are completed, based on the image data acquired by the image acquisition resources and the processing results of the image processing resources, a display image corresponding to the target display mode is generated and output to the display device.

[0055] For example, if the second processing task completes before the first processing task, the image acquisition resource remains active and continuously provides image data, waiting for the first processing task to complete before image generation and output. This ensures that the generated display image has both initialized image processing capabilities and available image data, avoiding the output of unprocessed raw images before image processing resources are ready, and guaranteeing that the quality of the output display image meets the requirements of the target display mode.

[0056] In this embodiment of the disclosure, by performing the initialization of image processing resources and the activation of image acquisition resources on which the target display mode depends at least partially in parallel in timing, the overall loading time required from the issuance of the trigger switching command to the final display image is effectively shortened, the response speed and operation smoothness of the display device when switching to a specific display mode are improved, the waiting time of the user during the mode switching process is reduced, the user interaction experience is improved, and no additional hardware cost is required.

[0057] In some embodiments, performing a first processing task and a second processing task includes: performing the initialization of a basic function software package corresponding to a target display mode in parallel; and performing the first processing task and the second processing task in parallel after the initialization of the basic function software package is completed.

[0058] The basic functional software package can be, for example, a software development kit (SDK) corresponding to the target display mode. For instance, taking night vision mode as the target display mode, the basic functional software package can be the infrared night vision algorithm functional package (i.e., thermal imaging SDK) of the infrared night vision system. This package is integrated into the vehicle's infotainment software to provide a basic operating environment for subsequent image processing resource initialization and image acquisition resource activation.

[0059] The above process can be understood as a hierarchical scheduling strategy: the initialization of the basic functional software package is completed first as a prerequisite, providing a basic runtime environment that the first and second processing tasks share a common dependency on. For example, this basic runtime environment includes loading the algorithm library, handshaking the hardware abstraction layer, allocating the memory pool, and defining data formats and preparing thread scheduling services. After the basic functional software package initialization is complete, the first and second processing tasks no longer have a mutual dependency, thus allowing them to be executed safely in parallel, avoiding parallel task execution failure due to the incomplete preparation of the basic runtime environment.

[0060] It is worth noting that in related technologies, the initialization of image processing resources is included within the initialization process of the basic function software package. This means that the initialization process of the basic function software package must sequentially execute the basic environment preparation and image processing resource initialization, such as configuring the image processing pipeline, loading filter operators, initializing the color correction matrix, and warming up the image enhancement engine. These operations are typically time-consuming. Furthermore, the activation of image acquisition resources, including powering on the image sensor, establishing the video transmission channel, and allocating the frame buffer, is set to only begin after the basic function software package initialization is fully completed. This sequential dependency causes the activation of image acquisition resources to be unnecessarily delayed, forming a sequential bottleneck in the switching process. The overall loading time is the sum of the time spent on basic environment preparation, image processing resource initialization, and image acquisition resource activation, significantly prolonging the switching loading time.

[0061] This embodiment of the disclosure separates the initialization of image processing resources from the basic function software package into an independent first processing task. This ensures that the initialization of image processing resources is no longer included in the initialization process of the basic function software package and is independent of the activation of image acquisition resources, which is the second processing task. This eliminates the serial blocking of image acquisition resource activation by image processing resource initialization, allowing the first and second processing tasks to be started simultaneously after the basic function software package initialization is completed. This reduces the total time originally accumulated serially to a shorter time after parallel execution. This shorter time is determined by the longer of the first and second processing tasks combined with the initialization time of the basic function software package.

[0062] In some embodiments, the start time of the first processing task is the same as the start time of the second processing task, that is, the two tasks start at the same time.

[0063] Compared to the approach where two tasks are started sequentially but partially in parallel, in this embodiment, the first and second processing tasks start at the same time, which can achieve maximum overlap between the two tasks on the timeline.

[0064] In a sequential but partially parallel approach, the time difference between the start times of the two tasks prevents their execution periods from being perfectly aligned, resulting in some serial latency that cannot be eliminated. Consequently, the total execution time still includes the additional waiting time corresponding to this time difference. This embodiment sets the start times of the first and second processing tasks to the same moment, allowing both tasks to enter a parallel execution state from the start. This eliminates the serial latency caused by the start time difference, thereby compressing the total time of mode switching into the single time of the longer task, achieving the optimal time compression effect.

[0065] For example, image processing resource initialization and image acquisition resource activation can be initiated simultaneously at the same time as the basic function package initialization is completed. For instance, in response to an indication signal indicating the completion of basic function package initialization, the execution of the first and second processing tasks can be triggered simultaneously. In implementation, independent execution threads can be created to respectively handle the first and second processing tasks, allowing both tasks to be scheduled and executed in parallel at the same time. For example, a first thread can be created to perform image processing resource initialization, and a second thread can be created to perform image acquisition resource activation; the first and second threads can be created simultaneously or simultaneously set to a runnable state to achieve simultaneous startup.

[0066] In this case, the total time for mode switching is equal to the sum of the initialization time of the basic function package and the longer of the first and second processing tasks, rather than the sum of the times of the two tasks or the time of partial superposition, thus further achieving time compression.

[0067] The execution flow of the second processing task will be further explained below.

[0068] In some embodiments, the second processing task includes: issuing an enable command to the image acquisition resource and obtaining image data fed back by the image acquisition resource; after confirming that the image data meets the target availability conditions, if the first processing task has not yet been completed, keeping the image acquisition resource in the enabled state and waiting for the first processing task to be completed.

[0069] When issuing a start command to an image acquisition resource, a preset communication protocol can be used to ensure that the start command is reliably transmitted to the image acquisition resource, enabling it to start up and enter the working state as expected. This communication protocol can be determined based on the hardware interface of the image acquisition resource. For example, taking a night vision camera as the image acquisition resource, a start command containing a power-on sequence and register configuration parameters can be issued to the night vision camera via the I2C protocol, an inter-integrated circuit bus in a low-voltage differential signal harness. After correctly parsing the command, the night vision camera executes its internal startup process.

[0070] After receiving the start command and completing the startup, the image acquisition resource begins acquiring images and feeding the image data back to the system. The system acquires the image data fed back by the image acquisition resource, such as receiving real-time transmitted video data frames through the video input interface, and caches the frame data in a preset storage area. The system can perform usability detection on the image data, including detecting the integrity of the image frames, frame rate stability, brightness distribution, etc., thereby providing a data basis for subsequent confirmation of whether the image data meets the target usability conditions.

[0071] After confirming that the image data meets the target availability conditions, if the first processing task has not yet been completed, the image acquisition resource remains enabled, and the system waits for the first processing task to complete. The target availability conditions characterize that the image acquisition resource has been successfully activated and that the quality of the output image data meets the input requirements for subsequent image processing. In some embodiments, the target availability conditions may include: continuously receiving a preset number of valid image data frames with the frame interval within a preset tolerance range; the average brightness value of the image data being higher than a preset dark field threshold; and the absence of invalid lines or invalid frame markers in the image data due to transmission errors.

[0072] When the second processing task completes before the first, the image acquisition resource remains on instead of being turned off and then on again. This avoids the power-on delay, register reconfiguration time, and additional time overhead of waiting for image data to stabilize caused by repeatedly starting the image acquisition resource. Simultaneously, while waiting for the first processing task to complete, the image acquisition resource continuously provides available image data and updates the cache. This ensures that once the first processing task completes, the latest available image frames can be immediately acquired and fed into the initialized image processing pipeline for image generation and output, achieving a seamless transition between image acquisition readiness and image processing startup.

[0073] The embodiments of this disclosure, through the above-described method, ensure that the image acquisition resources, once ready, are not shut down or restarted due to the incomplete completion of the first processing task, but remain active, waiting for the completion of the first processing task. Therefore, after both tasks are completed, the system simultaneously possesses initialized image processing resources and image acquisition resources that are active and outputting valid image data. It can immediately generate and output images based on the current image frame, avoiding the additional delay introduced by restarting the image acquisition resources and further shortening the overall time from mode switching to the presentation of the first image frame.

[0074] The above text explains that image data needs to meet the target usability conditions. The following text further explains how to confirm that image data meets the target usability conditions.

[0075] In some embodiments, confirming that the image data meets the target availability conditions includes: receiving a normal flag bit reported by the image acquisition resource, wherein the normal flag bit is used to indicate that the image acquisition resource has completed the shutter action and the output image data quality is normal; and confirming that the image data meets the target availability conditions based on the normal flag bit.

[0076] The normal flag is proactively reported by the image acquisition resource after the shutter action is completed and the quality of the output image data is confirmed to be normal. For example, taking a night vision camera as the image acquisition resource, after receiving an opening command, the night vision camera executes the shutter opening action. After the shutter opening action is completed and the internal diagnostic module determines that the output image data is normal, it reports the normal flag to the system through a preset communication interface. In some embodiments, this reporting process can be implemented based on an interrupt signal or a status register update, without requiring the system to initiate an active query.

[0077] The shutter action can be the opening of the shutter window of a night vision camera or the initialization and reset of an electronic shutter. The completion of the shutter action signifies that the optical path of the image acquisition resource has been opened, allowing the photosensitive element to normally receive light signals from the external scene and convert them into electrical signals, thereby generating valid image data. Using the completion of the shutter action as one of the criteria for determining the normal startup of the image acquisition resource ensures that the image acquisition resource only reports the normal flag after the optical path is fully open and the photoelectric conversion link is established. This avoids classifying image data as usable before the shutter is fully open or the photosensitive element is still in an unstable state, thus ensuring that the image data used for subsequent processing has sufficient signal-to-noise ratio and scene reproduction quality.

[0078] By proactively reporting a normal status flag from image acquisition resources, the system eliminates the need to poll image data frames to determine their quality or continuously monitor the status register of the image acquisition resources. This reduces the system's processing burden and communication bandwidth consumption, allowing the system to focus its computational resources on the primary processing task or other parallel operations while waiting for the image acquisition resources to become ready. Upon receiving the normal status flag, the system marks the image acquisition resources as ready and records the ready time. Subsequently, after the primary processing task is completed, the system can directly call the currently cached image data provided by the image acquisition resources for image generation and output, eliminating the additional delay caused by status confirmation.

[0079] Furthermore, the normal status flag serves as a confirmation signal for image data availability, providing a clear state synchronization mechanism for the system. The first and second processing tasks coordinate based on their respective status flags: the first processing task uses its own initialization completion flag as the readiness criterion, while the second processing task uses the received normal status flag as the readiness criterion; once both are ready, they enter the image generation stage. This mechanism ensures clear judgment criteria for coordination between the two tasks, avoiding startup timing deviations due to misjudgments of status or timeouts, and ensuring that the image generation stage can begin immediately upon completion of both tasks.

[0080] This disclosure provides a clear judgment mechanism based on the self-state feedback of the image acquisition resource, enabling the system to accurately determine whether the image acquisition resource has been started and output image data of normal quality, without relying on external detection methods or preset waiting time, thus improving the accuracy of state judgment and the timeliness of system response.

[0081] In some embodiments, the image display mode switching method further includes: preprocessing the image data fed back by the image acquisition resource before the first processing task has been completed.

[0082] Preprocessing can include basic image processing operations that do not rely on image processing resources, such as image denoising, image format conversion, image cropping, white balance correction, or color space conversion. Preprocessing the image data within the time window while waiting for the first processing task to complete allows some image processing tasks that would normally begin after both the first and second processing tasks are finished to be performed earlier.

[0083] Accordingly, based on the image data acquired by the image acquisition resources and the processing results of the image processing resources, a display image corresponding to the target display mode is generated, including: generating a display image based on the preprocessed image data and the processing results of the image processing resources.

[0084] Once the first processing task is completed, subsequent image generation can be performed directly based on the pre-processed image data, reducing the delay from the completion of both the first and second processing tasks to the final output and display image, and further improving the response speed of display mode switching.

[0085] In this embodiment, by shifting some image processing tasks forward, idle computing resources during the waiting period for the first processing task to complete can be fully utilized. After the first processing task is completed, the preprocessed image data is input to the initialized image processing resource for processing, and the final display image is generated based on the processing result. Since the image data has already been preprocessed, the image processing resource can directly perform core processing upon receiving the preprocessed image data, without needing to perform basic image preprocessing steps again. This shortens the time from receiving data to outputting processing results, making the generation of the final display image more efficient.

[0086] In some embodiments, the target display mode is night vision mode, the image processing resource is an artificial intelligence (AI) image recognition model, and the image acquisition resource is an infrared camera.

[0087] Night vision mode refers to the display device switching to a mode for displaying nighttime images. For example, it could be a night vision display mode based on infrared thermal imaging technology, used to provide drivers with a visual image of the road ahead in nighttime or low-visibility conditions. In night vision mode, infrared image data captured by an infrared camera is processed by an AI image recognition model and displayed on the screen as an enhanced night vision image with overlaid target markers. Night vision mode requires infrared image data captured by an infrared camera as input, and simultaneously relies on an AI image recognition model to perform target recognition and rendering processing on the infrared image to generate the enhanced image; both are necessary preparatory steps in the mode switching process. Executing the infrared camera's activation preparation process and the AI ​​image recognition model's initialization process in parallel can effectively shorten the overall waiting time from triggering the night vision mode switching command to displaying the first frame of night vision image, improving the driver's experience and responsiveness in nighttime driving scenarios.

[0088] AI image recognition models are used to identify and render the contours of targets in infrared images, thereby highlighting pedestrians, vehicles, animals, and other objects in night vision images. For example, AI image recognition models can be implemented based on deep convolutional neural network structures. Their initialization process includes model structure construction, pre-trained weight loading, inference engine instantiation, and model warm-up inference, which typically takes a considerable amount of time. Compared to traditional image processing algorithms based on threshold segmentation or feature matching, AI image recognition models can achieve more refined target recognition and contour rendering, such as upgrading from traditional rectangular target selection to pixel-level target contour rendering based on instance segmentation, making target boundaries more accurate and blending more naturally with the background. However, due to the more complex model structure and larger number of parameters, the initialization time also increases accordingly. By parallelizing the initialization of the AI ​​image recognition model with the activation of the infrared camera, the model initialization time and camera activation time can be overlapped without reducing image processing capabilities and recognition accuracy, thus shortening the overall time of mode switching.

[0089] Infrared cameras are used to collect infrared image data of the road scene ahead, providing real-time input data for AI image recognition models. The activation process of an infrared camera involves steps such as sending an activation command to the camera, the camera executing a shutter action, establishing a video data output channel, and stabilizing the output of valid image frames, all of which require a certain amount of time. This activation process is independent of the initialization of the AI ​​image recognition model in terms of resource usage and execution logic, with no serial dependencies at the hardware or software level. Executing the activation of the infrared camera and the initialization of the AI ​​image recognition model in parallel allows the infrared camera to complete power-on reset, shutter opening, and stable image data output during the AI ​​image recognition model initialization process. This avoids the additional delay of waiting separately for the infrared camera to activate after model initialization, thereby further reducing the time interval between mode triggering and the availability of image processing and display conditions.

[0090] In night vision mode, the first processing task includes: loading the model parameters of the artificial intelligence image recognition model; and initializing the inference and running environment of the artificial intelligence image recognition model on the vehicle chip.

[0091] The model parameters for loading an AI image recognition model can include weight parameters and network structure parameters. The weight parameters are numerical parameters obtained after model training, while the network structure parameters define the number of layers, layer types, connection relationships, and input / output dimensions. For example, the model parameters include parameters for the main model and parameters for smaller models. The main model performs instance-level segmentation and contour rendering of targets in infrared images, while the smaller models perform rapid classification of target categories or preprocessing and filtering of image regions. The main model and smaller models correspond to different recognition tasks or processing stages, and their parameters can be loaded sequentially or in parallel.

[0092] Loading model parameters is one of the most time-consuming steps in the initialization of AI image recognition models. This time consumption is even more pronounced when the model structure is complex. For example, models using target contour rendering have deeper network layers, a larger number of parameters, and correspondingly larger weight files compared to traditional target selection models, significantly increasing the time required to read and parse them from storage media into memory. By incorporating model parameter loading into a parallel first processing task, this operation overlaps with the activation of the infrared camera in the second processing task, preventing the model parameter loading process from blocking the activation of image acquisition resources and thus avoiding unnecessary delays in activating image acquisition resources due to the long loading time.

[0093] When initializing the inference runtime environment of the AI ​​image recognition model on the vehicle chip, the vehicle chip can be, for example, an on-board processing chip with neural network acceleration capabilities, such as an automotive-grade system-on-a-chip that integrates a digital signal processor, a neural network processing unit, or a graphics processor.

[0094] The inference runtime environment can be a runtime environment created based on a neural network inference framework for the vehicle's infotainment chip, including operations such as device creation, computation graph compilation, memory allocation, and the establishment of inference sessions. For example, the inference runtime environment can be a neural network-based inference runtime environment. The initialization of the inference runtime environment provides the necessary hardware abstraction and computational resource configuration for the actual inference execution of the AI ​​image recognition model on the vehicle's infotainment chip, ensuring that model weight parameters are correctly mapped to the accelerator memory space and that the computation graph is compiled into a hardware-executable instruction sequence. This initialization process involves computation graph compilation and hardware resource configuration, which is time-consuming. Integrating the initialization of the inference runtime environment and model parameter loading into the first processing task allows the first processing task to be executed in parallel with the second processing task, thereby reducing the overall loading latency from triggering the night vision mode to achieving full image processing capabilities through time overlap.

[0095] It is understood that the image display mode switching method provided in this disclosure is not limited to night vision mode application scenarios. Any display scenario that requires loading a large number of parameter models during display mode switching can adopt the parallel processing approach of this disclosure, which executes the initialization of image processing resources such as model loading and the activation of corresponding image acquisition resources in at least partial parallel time, thereby compressing the overall time consumption of mode switching.

[0096] For example, when switching to panoramic surround view mode, a deep learning model for multi-image stitching and bird's-eye view generation needs to be loaded. The initialization of this model can be performed in parallel with the simultaneous activation of multiple cameras. Similarly, when switching to intelligent scene recognition mode, a neural network model for environmental semantic understanding and scene classification needs to be loaded. The initialization of this model can be performed in parallel with the activation of the corresponding scene acquisition camera. Both scenarios face the problem of large model size and long loading time. The method disclosed herein is equally applicable, effectively shortening the waiting time during mode switching and improving the overall interactive smoothness of the display device and the user experience.

[0097] In some embodiments, the image display mode switching method is applied to a controller, which may be an in-vehicle control unit such as a vehicle infotainment controller, instrument cluster controller, or domain controller, or other control devices or computing platforms with image processing capabilities. The controller can uniformly manage the initialization of image processing resources, the activation of image acquisition resources, and the generation and output of display images, thereby achieving centralized control and optimized scheduling of the display mode switching process.

[0098] When the display device is in the current display mode, if the controller's processing resources meet the preset idle conditions, at least part of the initialization data of the image processing resources on which the target display mode depends is preloaded.

[0099] It is understandable that the determination and preloading process of the target display mode can be triggered in the following ways: preparatory operations such as users entering the switching interface or browsing switching options, changing trends of environmental parameters such as ambient brightness or rainfall intensity, predictions based on fixed time periods or specific geographical locations based on historical data, or system events such as changes in navigation routes.

[0100] After determining the target display mode and receiving the trigger signal for the preloading process, if the controller's processing resources meet the preset idle conditions, the idle processing resources can be used to preload at least a portion of the initialization data of the image processing resources on which the mode depends.

[0101] Preset idle conditions could include, for example, the controller's processor utilization being below a preset threshold, available memory being above a preset threshold, or no high-priority tasks currently being executed. At least partial initialization data could include, for example, model parameter files for AI image recognition models, configuration data for the inference framework, or compilation results of the computation graph.

[0102] By preloading initialization data when the controller is idle, the data loading process, which would normally only occur after a mode switch is triggered, is completed ahead of time. Since preloading occurs during the normal operation of the current display mode, it does not affect the performance of the current display mode, while also reserving initialization data for subsequent mode switches.

[0103] When executing the first processing task after the actual trigger mode switch, the initialization of image processing resources is completed directly using the initialization data that has been pre-loaded into memory or cache, without having to load all the initialization data from scratch.

[0104] In this embodiment of the disclosure, the workload of the first processing task is reduced during actual execution through the preloading process of initialization data, and the execution time is shortened accordingly. In the scenario where it is executed in parallel with the second processing task, the shortening of the time of the first processing task means that the time when both parallel tasks are completed is brought forward, thereby further compressing the total time of display mode switching and making the mode switching delay perceived by the user shorter.

[0105] Figure 3 The diagram shows a timing flow of an image display mode switching method in related technologies. The mode switching process in related technologies is executed serially. For example... Figure 3As shown, during the system initialization and preparation phase, the vehicle infotainment software initiates an SDK initialization call, which simultaneously triggers and executes AI model initialization. In this phase, AI model initialization is an independent activity within the SDK initialization process. Its execution includes time-consuming operations such as model loading, and subsequent steps are constrained by condition A: the SDK initialization activity must complete AI model initialization before it can continue. Therefore, the SDK initialization phase includes a sequential waiting process for AI model initialization. Only after model initialization is complete does the SDK initialization return the result to the vehicle infotainment software, thus completing all the preparatory work for the system's core functional components.

[0106] Next, the system enters the trigger startup and image processing phase. Specifically, the system sends a signal to trigger the activation of the night vision system. This signal sends an activation command to the camera, initiating its hardware startup process. Simultaneously, the image processing module receives image data and enters a waiting state under the logical constraint of condition B. In other words, the image processing module needs to wait for the camera's night vision function to be fully activated before proceeding with subsequent judgment and processing. This means that the camera's hardware setup process and the image processing startup logic also form a serial dependency.

[0107] Once night vision is activated, the system recognizes that the night vision function is in place. The image processing module then reads and processes the transmitted image data, performs a normal image output judgment, and sends the normal image data that has passed the judgment to the vehicle software through the reporting process, thereby completing the switching and imaging of the entire night vision display mode.

[0108] Based on the above-mentioned timing interactions, it can be seen that after the relevant technology triggers the night vision mode, it needs to go through the synchronous waiting for AI model initialization, the waiting for the night vision camera to turn on, and finally the activation of image processing. The serial execution mechanism between the above-mentioned stages will cause the total image loading time to accumulate segment by segment. Especially when the model initialization takes a long time, the total image loading time will be significantly extended, thus directly affecting the driver's user experience.

[0109] Figure 4 The diagram shown is a timing flowchart of an image display mode switching method according to an embodiment of this disclosure. Figure 4 As shown, when the system receives a trigger signal to switch from the current display mode to night vision mode, it will execute the following steps.

[0110] S41, SDK Initialization. In this stage, the thermal imaging SDK integrated into the vehicle's infotainment software begins loading and prepares its own runtime environment, providing a unified foundation for the parallel execution of subsequent tasks. The initialization of this SDK no longer involves the simultaneous initialization of the AI ​​model, which has been separated into an independent task.

[0111] After the SDK initialization is complete, execute the following simultaneously: S42, start the first processing task; S43, start the second processing task.

[0112] The first processing task is to initialize the AI ​​image recognition model upon which night vision mode relies. This includes loading the model's weight parameters and network structure parameters, such as the main model parameters and sub-model parameters, and initializing the inference runtime environment based on the Qualcomm neural network framework on the vehicle's infotainment chip, completing operations such as device creation, computation graph compilation, and memory allocation. The second processing task is to activate the infrared camera upon which night vision mode relies. The switching device sends an activation command to the infrared camera via the I2C protocol in the low-voltage differential signal harness and acquires the image data fed back by the camera.

[0113] Upon receiving the activation command, the infrared camera opens its shutter window, opening the optical path and beginning to acquire infrared image data of the scene in front. When the camera confirms the shutter action is complete and the output image data quality is normal, it proactively reports a normal flag to the system. The system receives this flag and determines that the image data meets the target usability criteria. At this point, because the AI ​​model initialization is not yet complete, the system keeps the infrared camera on and waits for the first processing task to complete. The camera continues to provide image data, but the system does not generate a display image yet, avoiding outputting an unprocessed raw image before the AI ​​model is ready.

[0114] After a period of parallel processing, the AI ​​model's parameters were loaded and the inference environment was initialized. At this point, both the first and second processing tasks had been completed, and the system detected that both conditions were met. Subsequently, the switching device invoked the initialized AI model to perform target recognition and contour rendering on the image data collected by the infrared camera, generating a display image corresponding to the night vision mode, and outputting it to the instrument panel display. The display showed a night vision image of the road ahead with contour rendering, clearly identifying pedestrians, vehicles, and other targets.

[0115] In this scenario, the present invention significantly improves the response speed of night vision mode switching. Previously, the sequential process of SDK initialization, AI model initialization, and camera activation took up to nearly one second, causing a noticeable pause on the instrument panel after the driver pressed the button, impacting driving safety. With this solution, model initialization and camera activation proceed in parallel after SDK initialization, reducing the total switching time to approximately 390 milliseconds. The driver experiences almost no waiting time, resulting in a significant improvement in nighttime driving safety and operational smoothness.

[0116] The above text combined Figures 1 to 4 The method embodiments of this disclosure have been described in detail below, in conjunction with... Figure 5The apparatus embodiments of this disclosure are described in detail below. It should be understood that the descriptions of the method embodiments correspond to the descriptions of the apparatus embodiments; therefore, any parts not described in detail can be referred to the foregoing method embodiments.

[0117] Figure 5 The diagram shown is a structural schematic of an image display mode switching device provided in an embodiment of this disclosure. Figure 5 As shown, the image display mode switching device includes an execution module 510 and a display module 520.

[0118] The execution module 510 is configured to execute a first processing task and a second processing task in response to a trigger signal that switches the display device from the current display mode to the target display mode. The first processing task is to initialize the image processing resources on which the target display mode depends, and the second processing task is to enable the image acquisition resources on which the target display mode depends. The first processing task and the second processing task are executed in at least partially parallel time.

[0119] The display module 520 is used to generate a display image corresponding to the target display mode and output it to the display device based on the image data acquired by the image acquisition resource and the processing result of the image processing resource after both the first processing task and the second processing task have been completed.

[0120] In some embodiments, the execution module 510 is further configured to initialize the basic function software package corresponding to the target display mode; and after the initialization of the basic function software package is completed, execute the first processing task and the second processing task.

[0121] In some embodiments, the execution module 510 is further configured to issue an enable command to the image acquisition resource and obtain image data fed back by the image acquisition resource; after confirming that the image data meets the target availability conditions, if the first processing task has not yet been completed, the image acquisition resource is kept in the enabled state and the first processing task is waited for to be completed.

[0122] In some embodiments, the execution module 510 is further configured to receive a normal flag bit reported by the image acquisition resource, wherein the normal flag bit is used to indicate that the image acquisition resource has completed the shutter action and the output image data quality is normal; and based on the normal flag bit, confirm that the image data meets the target availability conditions.

[0123] In some embodiments, the execution module 510 is further configured to load the model parameters of the artificial intelligence image recognition model and initialize the inference runtime environment of the artificial intelligence image recognition model on the vehicle chip.

[0124] Below, for reference Figure 6 To describe a vehicle according to an embodiment of this disclosure. Figure 6 The diagram shown is a structural schematic of a vehicle provided according to an embodiment of this disclosure. Figure 6 As shown, vehicle 60 includes one or more processors 601 and memory 602.

[0125] The processor 601 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the vehicle 60 to perform desired functions.

[0126] The memory 602 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 601 may execute the program instructions to implement the image display mode switching methods of the various embodiments of this disclosure described above and / or other desired functions. The computer-readable storage medium may also store various contents such as port allocation algorithms, load balancer resource instances, development machine resource instances, etc.

[0127] In one example, vehicle 60 may also include input device 603 and output device 604, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0128] The input device 603 may include, for example, a touch screen, a keyboard, a mouse, etc. The output device 904 may include, for example, a display device, a speaker, and a communication network and its connected remote output devices, etc.

[0129] Of course, for the sake of simplicity, Figure 6 Only some of the components of the vehicle 60 relevant to this disclosure are shown, omitting components such as buses, input / output interfaces, etc. In addition, the vehicle 60 may include any other suitable components depending on the specific application.

[0130] In addition to the methods and apparatus described above, embodiments of this disclosure may also be computer program products, including computer program instructions that, when run by an image display mode switching system, cause the image display mode switching system to perform the steps in the image display mode switching methods according to various embodiments of this disclosure described above.

[0131] Computer program products can be written in any combination of one or more programming languages ​​to perform the operations of embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0132] Furthermore, embodiments of this disclosure may also be computer-readable storage media storing computer program instructions that, when run by the image display mode switching system, cause the image display mode switching system to perform the steps in the image display mode switching methods according to various embodiments of this disclosure described above.

[0133] Computer-readable storage media may take the form of any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0134] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0135] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A method for switching image display modes, characterized in that, include: In response to a trigger signal that switches the display device from the current display mode to the target display mode, a first processing task and a second processing task are executed in parallel. The first processing task is to initialize the image processing resources on which the target display mode depends, and the second processing task is to enable the image acquisition resources on which the target display mode depends. The first processing task and the second processing task overlap at least partially in timing. After both the first processing task and the second processing task are completed, a display image corresponding to the target display mode is generated and output to the display device based on the image data acquired by the image acquisition resource and the processing result of the image processing resource.

2. The image display mode switching method according to claim 1, characterized in that, The parallel execution of the first processing task and the second processing task includes: Execute the initialization of the basic function software package corresponding to the target display mode; After the basic functional software package is initialized, the first processing task and the second processing task are executed in parallel.

3. The image display mode switching method according to claim 1, characterized in that, Performing the second processing task includes: Send an enable command to the image acquisition resource and obtain the image data fed back by the image acquisition resource; After confirming that the image data meets the target availability conditions, if the first processing task has not yet been completed, the image acquisition resource remains in the enabled state, and the process waits for the first processing task to be completed.

4. The image display mode switching method according to claim 3, characterized in that, The confirmation that the image data meets the target usability conditions includes: Receive the normal flag bit reported by the image acquisition resource, wherein the normal flag bit is used to indicate that the image acquisition resource has completed the shutter action and the output image data quality is normal; Based on the normal flag bit, it is confirmed that the image data meets the target usability conditions.

5. The image display mode switching method according to claim 3, characterized in that, Also includes: Before the first processing task is completed, the image data fed back by the image acquisition resource is preprocessed; The step of generating a display image corresponding to the target display mode based on the image data acquired by the image acquisition resources and the processing results of the image processing resources includes: The display image is generated based on the preprocessed image data and the processing results of the image processing resources.

6. The image display mode switching method according to any one of claims 1 to 5, characterized in that, The target display mode is night vision mode, the image processing resource is an artificial intelligence image recognition model, and the image acquisition resource is an infrared camera.

7. The image display mode switching method according to claim 6, characterized in that, Executing the first processing task includes: Load the model parameters of the artificial intelligence image recognition model; Initialize the inference and runtime environment of the artificial intelligence image recognition model on the vehicle's infotainment chip.

8. The image display mode switching method according to any one of claims 1 to 5, characterized in that, The start time of the first processing task is the same as the start time of the second processing task.

9. The image display mode switching method according to any one of claims 1 to 5, characterized in that, Applied to controllers; Prior to the trigger signal that switches the display device from the current display mode to the target display mode, the method further includes: When the display device is in the current display mode, if the processing resources of the controller meet the preset idle conditions, at least part of the initialization data of the image processing resources on which the target display mode depends is preloaded. The execution of the first processing task includes: The initialization of the image processing resources is completed based on the pre-loaded initialization data.

10. A vehicle, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the display mode switching method according to any one of claims 1 to 9.

Images