Tire gap foreign object recognition method, vehicle, and storage medium

Abstract

The application relates to a tire gap foreign matter identification method, a vehicle and a storage medium, and the tire gap foreign matter identification method comprises the following steps: performing preliminary detection on a tire gap through a first detection unit to determine whether an abnormal area exists; in the case that the tire gap has the abnormal area, scanning the abnormal area through a second detection unit to identify foreign matters in the abnormal area; wherein the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit. According to the method, the tire gap is preliminarily detected through the relatively low-frequency first detection unit to determine whether the abnormal area exists in the tire gap, then the abnormal area existing in the tire gap is scanned through the relatively high-frequency second detection unit to identify the foreign matters in the abnormal area, the detection success rate of small-particle foreign matters in the tire gap can be improved, and then the safety hidden danger caused by the foreign matters embedded in the tire gap can be reduced.

Description

Technical Field

[0001] This application relates to the field of tire inspection technology, and in particular to a method for identifying foreign objects in tire gaps, a vehicle, and a storage medium. Background Technology

[0002] During vehicle operation, foreign objects such as stones can get stuck in the tire treads, affecting tire grip and even braking performance. Therefore, accurately identifying foreign objects in tire treads has become an urgent problem to be solved. Summary of the Invention

[0003] This application provides a method for identifying foreign objects in tire crevices, a vehicle, and a storage medium to solve the technical problem of how to accurately identify foreign objects in tire crevices.

[0004] In a first aspect, this application provides a method for identifying foreign objects in tire gaps, the method comprising: The first detection unit performs a preliminary inspection of the tire gaps to determine if any abnormal areas exist. In the event of an abnormal area in the tire tread, a second detection unit scans the abnormal area to identify foreign objects within it; wherein, the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit.

[0005] Optionally, the first detection unit performs a preliminary inspection of the tire gaps to determine if any abnormal areas exist, including: Continuously monitor the current vehicle speed; When the current vehicle speed is less than a preset threshold, the first detection unit performs a preliminary detection of the tire gaps to determine whether there are any abnormal areas.

[0006] Optionally, the first detection unit includes multiple first sensors disposed on the wheel arch; the first detection unit performs preliminary detection of tire gaps to determine whether abnormal areas exist, including: The first sensor collects an echo signal from the tire gap once at preset rotation intervals; All echo signals are beamformed to obtain an image of the inside of the trench; Determine whether there are abnormal areas in the image inside the trench.

[0007] Optionally, all echo signals are beamformed to obtain an image of the trench interior, including: Each target echo signal collected by the first sensor is aligned to obtain an alignment signal; The alignment signals from all the first sensors are beamformed to obtain an image of the inside of the trench.

[0008] Optionally, before aligning each of the target echo signals acquired by the first sensor, the method further includes: Obtain the preset tire tread groove feature model; The original echo signal collected by the first sensor is filtered according to the tire tread groove feature model to obtain a clean echo signal; The foreign object features in the pure echo signal are enhanced to obtain the processed target echo signal.

[0009] Optionally, feature enhancement is performed on the foreign object features in the clean echo signal to obtain the processed target echo signal, including: Obtain the preset enhancement coefficient; The foreign object feature is enhanced based on the tire tread groove feature model, the enhancement coefficient, and the pure echo signal to obtain the processed target echo signal.

[0010] Optionally, if an abnormal area exists in the tire tread, a second detection unit scans the abnormal area to identify foreign objects within it, including: In the event of an abnormal area in the tire tread, the abnormal area is scanned by the second detection unit to obtain the target scanning signal; The target scanning signal is input into a preset foreign object identification model to identify foreign objects in the abnormal area.

[0011] Optionally, after identifying foreign objects in the abnormal area, the method further includes: Obtain the recognition confidence level of the foreign object; When the identification confidence level is less than a first threshold, the abnormal region is marked so that it can be re-scanned during the next scan. When the identification confidence level is greater than or equal to the first threshold and less than the second threshold, a first prompt message is output; wherein, the first prompt message is used to indicate that there is a foreign object in the tire gap; When the identification confidence level is greater than or equal to the second threshold, a second prompt message is output; wherein, the second prompt message is used to indicate that there is a foreign object in the abnormal area.

[0012] Secondly, this application provides a foreign object identification device for tire gaps, the device comprising: The area recognition module is used to perform preliminary detection of tire gaps through the first detection unit to determine whether there are abnormal areas; The foreign object identification module is used to identify foreign objects in the abnormal area by scanning the abnormal area through a second detection unit when there is an abnormal area in the tire gap; wherein, the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit.

[0013] Thirdly, this application provides a vehicle including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When the processor executes a program stored in the memory, it implements the method for identifying foreign objects in tire gaps as described in any embodiment of the first aspect.

[0014] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for identifying foreign objects in tire gaps as described in any embodiment of the first aspect.

[0015] Compared with the prior art, the above-mentioned technical solution provided in this application has the following advantages: The method provided in this application performs preliminary detection of tire gaps by a first detection unit to determine whether there are abnormal areas; if there are abnormal areas in the tire gaps, the second detection unit scans the abnormal areas to identify foreign objects in the abnormal areas; wherein, the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit.

[0016] This method first uses a relatively low-frequency first detection unit to perform a preliminary inspection of the tire treads. Low-frequency ultrasound is used to determine if any abnormal areas exist within the treads. Because low-frequency ultrasound has strong penetrating power and a fast scanning speed, it can locate not only shallow but also deep abnormal areas containing foreign objects, and the localization speed is fast. This allows for a rapid and comprehensive inspection of the tire treads, and subsequent identification only requires identifying the abnormal areas. Then, a relatively high-frequency second detection unit performs a fine scan of the abnormal areas within the tire treads. High-frequency ultrasound is used to identify foreign objects within these areas, such as their shape or outline. Compared to low-frequency ultrasound, although the scanning speed of high-frequency ultrasound is relatively slower, its higher resolution and identification accuracy result in more precise identification of the shape or outline of the foreign objects.

[0017] Therefore, by combining low-frequency ultrasonic scanning and high-frequency ultrasonic scanning, low-frequency ultrasonic scanning can quickly identify abnormal areas. Then, during high-frequency ultrasonic scanning, only abnormal areas can be accurately identified without scanning non-abnormal areas. This reduces the area to be scanned by high-frequency ultrasonic scanning, improves identification efficiency, and increases the detection speed and success rate of small foreign objects in tire gaps. This can reduce the safety hazards caused by foreign objects embedded in tire gaps. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0021] Figure 1 A system architecture diagram of a method for identifying foreign objects in tire gaps, provided in one embodiment of this application; Figure 2 A flowchart illustrating a method for identifying foreign objects in tire gaps, provided as an embodiment of this application; Figure 3 A schematic diagram of a foreign object identification device in tire crevices provided in one embodiment of this application; Figure 4 This is a schematic diagram of the structure of a vehicle provided in one embodiment of this application. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] The following disclosure provides numerous different embodiments or examples for implementing various structures of this application. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of this application. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed.

[0024] To address the technical problem of accurately identifying foreign objects in tire treads in existing technologies, this application provides a method, vehicle, and storage medium for identifying foreign objects in tire treads. The method first uses a relatively low-frequency detection unit to perform a preliminary inspection of the tire treads, utilizing low-frequency ultrasonic waves to determine the presence of abnormal areas. Because low-frequency ultrasonic waves have strong penetrating power and fast scanning speed, they can locate not only shallow but also deep abnormal areas of foreign objects within the tire treads, and the location speed is fast, thus quickly completing a comprehensive inspection of the tire treads. Subsequent identification processes only require identifying the abnormal areas. Then, a relatively high-frequency detection unit performs a fine scan of the abnormal areas in the tire treads, using high-frequency ultrasonic waves to identify foreign objects within these areas, such as their shape or outline. Compared to low-frequency ultrasonic waves, although the scanning speed of high-frequency ultrasonic waves is relatively slower, its higher resolution and identification accuracy result in more precise identification of the shape or outline of the foreign objects. Therefore, by combining low-frequency ultrasonic scanning and high-frequency ultrasonic scanning, low-frequency ultrasonic scanning can quickly identify abnormal areas. Then, during high-frequency ultrasonic scanning, only abnormal areas can be accurately identified without scanning non-abnormal areas. This reduces the area to be scanned by high-frequency ultrasonic scanning, improves identification efficiency, and increases the detection speed and success rate of small foreign objects in tire gaps. This can reduce the safety hazards caused by foreign objects embedded in tire gaps.

[0025] The first embodiment of this application provides a method for identifying foreign objects in tire gaps. This method can be applied to, for example... Figure 1The system architecture shown includes at least a data acquisition module 101 and a data processing module 102, which establish a communication connection. The data acquisition module 101 may include a first detection unit and a second detection unit. The first detection unit may be a circumferential ultrasonic array disposed on the wheel arch, including multiple first sensors, such as eight 400-500kHz first sensors. The eight first sensors may be circumferentially distributed on the wheel arch above the tire, and the first sensors may be evenly distributed, for example, with each pair of first sensors spaced 45° apart to form a circle. The second detection unit may be a second sensor also disposed on the wheel arch above the tire, such as two 1MHz ultrasonic sensors. Specifically, this system architecture may be a tire tread foreign object detection system, or a vehicle equipped with a tire tread foreign object detection system. The type of vehicle is not limited, and may include, for example, a gasoline-powered vehicle, a pure electric vehicle, a hybrid electric vehicle, or a fuel cell vehicle, etc.

[0026] Next, based on this system architecture, the method for identifying foreign objects in tire gaps will be described in detail, such as... Figure 2 The method for identifying foreign objects in tire gaps includes: Step 201: The first detection unit performs a preliminary inspection of the tire gaps to determine whether there are any abnormal areas.

[0027] The first detection unit can be a circumferential ultrasonic array disposed on the wheel arch, including multiple first sensors, such as eight 400-500kHz first sensors. These eight first sensors can be circumferentially distributed on the wheel arch above the tire, and can be evenly distributed, for example, with each pair of first sensors spaced 45° apart to form a circle. The wavelength of the first sensors can be approximately 3-3.75mm. Due to its low frequency, the ultrasonic waves have low attenuation and strong penetration, allowing it to detect the bottom of the grooves and penetrate deep into the grooves to locate abnormal areas. Furthermore, the scanning speed is fast, enabling the location of both shallow and deep abnormal areas containing foreign objects in the tire treads. The pulse width can be 20μs. By performing preliminary detection of the tire treads using the relatively low-frequency first detection unit, the presence of abnormal areas can be determined, thereby narrowing down the detection range of the relatively high-frequency second detection unit.

[0028] In one embodiment, the first detection unit performs a preliminary detection of the tire gap to determine whether there is an abnormal area, including: continuously monitoring the current vehicle speed; when the current vehicle speed is less than a preset threshold, the first detection unit performs a preliminary detection of the tire gap to determine whether there is an abnormal area.

[0029] In this embodiment, to ensure accurate identification, the tire rotation should not be too fast. Therefore, the current vehicle speed can be continuously monitored, and the foreign object identification process in the tire gap is only initiated when the vehicle speed is lower than a preset threshold. The preset threshold can be set to, for example, 5 km / h or other relatively low speeds. This can be performed when the vehicle is decelerating to a near stop, or when the vehicle is starting to move, as long as the current vehicle speed is lower than the preset threshold.

[0030] Under high-speed conditions, the noise component of the echo signal received by the sensor increases dramatically, the inherent reflection waveform of the groove becomes disordered, and weak abnormal signals such as stones and cracks are easily masked by noise, making it impossible to extract clear features and accurately identify abnormal areas. Therefore, in this embodiment, the current vehicle speed is continuously monitored, and the foreign object detection process in the tire groove is only initiated when the vehicle speed is lower than a preset threshold. The preset threshold can be set to, for example, 5 km / h or other relatively low vehicle speeds. The initial tire groove detection process of the first detection unit is only initiated when the real-time monitored current vehicle speed value is lower than the preset threshold. Once the vehicle speed exceeds the preset threshold, the scanning and detection action of the first detection unit is automatically suspended to avoid invalid detection and misjudgment, ensuring that all collected signal data has analytical value.

[0031] In one embodiment, the first detection unit performs preliminary detection on the tire groove to determine whether there is an abnormal area, including: the first sensor collects an echo signal from the tire groove once at a preset rotation step interval; beamforming all the echo signals to obtain an image inside the groove; and determining whether there is an abnormal area in the image inside the groove.

[0032] In this embodiment, to address the limitation of the physical aperture of a static multi-sensor array, which makes it difficult to achieve the resolution required for detecting small particles, an echo signal can be collected from the tire groove at preset rotation step intervals to increase the virtual aperture. For example, a complete tire rotation is 360°, and the preset rotation step can be set to 15°, meaning echo signal data is collected every 15°. This allows for the collection of 24 sets of echo signal data during a tire rotation, signifying the acquisition of observation data from 24 different locations. The effective aperture length of the virtual aperture = physical sensor spacing × (360° / preset rotation step). Therefore, by rotating the tire, each first sensor can collect echo signal data at different angles, ultimately equivalent to an array of 24 virtual sensors, effectively increasing the effective aperture of the static sensors by 24 times. Subsequently, all echo signals are beamformed to obtain an image of the groove interior, which is then used to determine if any abnormal regions exist within the groove image. In this embodiment, no mechanical moving parts are required. The virtual aperture is increased by utilizing the natural rotation of the tires when the vehicle is moving at low speed or turning slightly in place, thereby improving the detection capability. The system structure is simple and reliable.

[0033] In one embodiment, beamforming all echo signals to obtain an image inside the trench includes: aligning the target echo signals collected by each first sensor to obtain an alignment signal; and beamforming all the alignment signals of the first sensors to obtain an image inside the trench.

[0034] In this embodiment, the tire angle corresponding to each collected target echo signal can be recorded. The target echo signals collected from different angles are aligned according to their arrival time to obtain an aligned signal. Then, the Delayed Summation (DAS) algorithm is used for beamforming to obtain an image of the inside of the trench.

[0035] Specifically, the mathematical expression for beamforming can be: I(x, y)=Σ[s_i(t-τ_i(x, y))×w_i(θ)] Where s_i is the echo signal acquired in the i-th acquisition; θ is the tire angle when the i-th echo signal is acquired; τ_i(x,y) is the propagation time from point (x,y) to the first sensor; t-τ_i(x,y) represents aligning the echo signals at different times; w_i(θ) is a weighting function based on the tire angle, that is, weighting the signal according to the tire angle; I(x,y) represents the preliminary detection result of the position of (x,y) inside the tire after beamforming.

[0036] In one embodiment, before aligning the target echo signals collected by each first sensor, the method further includes: acquiring a preset tire tread groove feature model; filtering the original echo signals collected by the first sensors according to the tire tread groove feature model to obtain a clean echo signal; and performing feature enhancement on the foreign object features in the clean echo signal to obtain the processed target echo signal.

[0037] In this embodiment, since the tire tread groove structure itself generates strong reflections, it may mask the weak signals of foreign objects such as stone particles. A tire tread groove feature model can be preset, and the original echo signal collected by the first sensor can be filtered by the tire tread groove feature model to obtain a pure echo signal. Furthermore, in order to improve the foreign object characteristics, the foreign object characteristics in the pure echo signal can be enhanced to obtain the processed target echo signal.

[0038] In this embodiment, the preset tire tread groove feature model can be: h groove(x)=A×exp(-|x| / λ)×cos(2πf×x+φ) Where A represents the amplitude of the groove echo reflected by the groove itself, λ represents the spatial attenuation coefficient of the groove echo, f represents the spatial frequency of the repeating arrangement of the tire tread grooves, and φ represents the initial phase of the groove echo waveform. The original echo signal collected by the first sensor is filtered using the tire tread groove characteristic model to obtain a clean echo signal. The interference of the original echo signal can be filtered as follows: s_clean = s_original - α × h_groove(x), where s_clean represents the clean echo signal, s_original represents the original echo signal, and α is an adaptive coefficient that can be dynamically adjusted according to the real-time groove echo signal. For example, older tires have shallower grooves and weaker reflections, so the adaptive coefficient α can be taken as a relatively small value; newer tires have deeper grooves and stronger reflections, so the adaptive coefficient α can be taken as a relatively large value. If α is a fixed coefficient, groove interference may not be filtered cleanly, or over-filtering may filter out the effective echo of foreign objects.

[0039] In one embodiment, feature enhancement is performed on foreign object features in the pure echo signal to obtain a processed target echo signal, including: obtaining a preset enhancement coefficient; and performing feature enhancement on the foreign object features based on the tire tread groove feature model, the enhancement coefficient, and the pure echo signal to obtain the processed target echo signal.

[0040] In this embodiment, to further improve the accuracy of foreign object identification, feature enhancement can be applied to the foreign object characteristics. For example, it can be enhanced using the following formula: F_foreign_matter(f,t) = |S_TFT(s_pure)| 2 -β×|STFT(h trench)| Where: STFT is the short-time Fourier transform, and β is the enhancement coefficient. Through time-frequency domain quadratic difference processing, residual interference components in the trench are further eliminated, relatively enhancing the foreign object characteristics within the trench, thus facilitating subsequent foreign object identification.

[0041] In this embodiment, by pre-setting the tire tread groove feature model, the anti-interference ability can be improved, and the stones can be effectively distinguished from mud, water droplets, etc., significantly reducing the false alarm rate.

[0042] Specifically, in the above embodiments of this application, the first detection unit is used to complete the preliminary detection of the tire groove gap. The purpose of the preliminary detection is to use a relatively low-frequency ultrasonic sensor to quickly locate the abnormal area in the tire groove gap where foreign objects are attached, laying the foundation for subsequent refined detection.

[0043] The initial detection process utilizes the vehicle's driving status to determine start-stop conditions. This process is executed under stable low-speed conditions to ensure the effectiveness and rationality of the initial detection. At high speeds, the noise components in the echo signals received by the sensors increase dramatically, the inherent reflection waveforms of the grooves become disordered, and weak abnormal signals such as stones and cracks are easily masked by noise, making clear feature extraction impossible and accurate identification of abnormal areas difficult. Therefore, in the embodiments described above, the current vehicle speed is continuously monitored. The foreign object identification process in the tire treads is only initiated when the vehicle speed is below a preset threshold, such as 5 km / h or other relatively low speeds. The initial tire tread detection process of the first detection unit is only initiated when the real-time monitored current vehicle speed is below the preset threshold. Once the vehicle speed exceeds the preset threshold, the scanning and detection actions of the first detection unit are automatically paused to avoid invalid or misjudged detections, ensuring that all collected signal data has analytical value.

[0044] Preliminary detection can be performed when the vehicle is decelerating to near a stop, or when the vehicle is starting to move, as long as the current vehicle speed is less than a preset threshold. Once the low-speed detection conditions are met, the first detection unit can be immediately activated to execute the preliminary detection process. In this embodiment, the first detection unit refers to eight sets of 400kHz-500kHz ultrasonic sensors arranged evenly around the circumference directly above the tire. The angle between adjacent ultrasonic sensors is controlled within 45°±5°, forming a complete circumferential scanning array that can cover the entire circumference of the tire tread and grooves without blind spots. When the vehicle speed is less than the preset threshold, for example, when the tire rotates at a low speed of 0.5 rpm, the tire rotates smoothly and gently, and the tread, grooves, and sidewall structure remain stable, without significant stretching or compression deformation. Combined with the relatively fixed ultrasonic emission and reflection angles of the first detection unit, the echo signal waveform is regular and stable, greatly reducing detection interference caused by motion deformation.

[0045] Eight sets of circular array sensors simultaneously emit ultrasonic detection signals. The sound waves cover all grooves and gaps on the tire surface vertically and obliquely. After contacting the tread structure, the inner walls of the grooves, and any adhering objects within the grooves, they form reflected echoes. The ultrasonic sensors simultaneously capture all the original echo signals, completing the data acquisition of the entire tire groove area. The acquired original echo signals are complex in composition. They include high-intensity fixed reflection signals generated by the tire tread groove structure itself, as well as weak reflection signals from abnormal targets such as stones, cracks, and detached rubber pieces inside the grooves. They may also include random noise signals such as driving vibrations. The various signals in the directly acquired original echo signals are superimposed and mixed, resulting in interference, making it impossible to accurately determine the tire condition based on the original echo signals.

[0046] To address the interference issues in the original echo signal, preprocessing can be performed. First, groove feature modeling is performed, for example, building a realistic tire tread groove feature model: h_groove(x) = A × exp(-|x| / λ) × cos(2πf × x + φ), where A represents the amplitude of the groove echo reflected by the groove itself, λ represents the spatial attenuation coefficient of the groove echo, f represents the spatial frequency of the repeating tire tread grooves, and φ represents the initial phase of the groove echo waveform. The original echo signal acquired by the first sensor is then filtered using the tire tread groove feature model to obtain a clean echo signal. Interference in the original echo signal can be filtered as follows: s_clean = s_original - α × h_groove(x), where s_clean represents the clean echo signal, s_original represents the original echo signal, and α is an adaptive coefficient that can be dynamically adjusted based on the real-time groove echo signal. For example, when dealing with old tires that are severely worn and have shallow grooves, the groove reflection signal is generally weak. A relatively small α value is automatically selected to slightly cancel out the groove model signal, preventing excessive filtering that reduces the effective echo of foreign objects within the groove. Conversely, for brand-new tires with intact grooves and sufficient depth, the groove reflection intensity is high. A relatively large α value is automatically selected to fully filter out high-intensity groove background interference. After adaptive filtering, most of the inherent groove reflection components are eliminated, resulting in a clean echo signal with prominent impurities and defects.

[0047] After completing the time-domain filtering, to further improve the accuracy of foreign object identification, feature enhancement can be applied to the foreign object features. For example, it can be enhanced using the following formula: F_foreign_matter(f,t) = |S_TFT(s_pure)| 2 -β×|STFT(h trench)| Where: STFT is Short-Time Fourier Transform, and β is the enhancement coefficient. Through time-frequency domain quadratic difference processing, residual interference components in the groove are further eliminated, relatively enhancing the characteristics of foreign objects within the groove, making subsequent foreign object identification easier. The enhancement coefficient β is used to proportionally reduce the time-frequency characteristics of the standard groove, further eliminating residual groove clutter in the time-frequency dimension, relatively amplifying the signal energy of unconventional structures such as stones and damage, widening the characteristic difference between normal tread and abnormal targets, weakening the ineffective interference of lightweight attachments such as mud and water droplets, and clearly revealing various abnormal features hidden within the groove gaps.

[0048] After obtaining the processed target echo signal, the system can determine whether there are abnormal areas in the tire grooves. For example, the real-time extracted target echo signal can be matched and verified against a pre-set database of normal tire groove features, comparing indicators such as signal energy strength, waveform variation patterns, and time-frequency distribution characteristics. When the signal characteristics of a local area deviate from the normal standard range, exhibiting phenomena such as sudden energy changes, waveform distortion, or abnormal spectral peaks, the system can determine that the location is an abnormal area in the tire groove. At this point, basic information such as the tire circumference angle, approximate area, and degree of signal abnormality corresponding to the abnormal area can be recorded simultaneously, completing the initial detection and judgment work.

[0049] It should be understood that if the initial inspection does not identify any abnormal areas, the vehicle speed monitoring and the first detection unit can remain in standby scanning state, waiting for the next low-speed condition inspection; once the inspection determines that there are abnormal areas in the tire groove gaps, the subsequent further inspection process for the abnormal areas will be carried out.

[0050] Step 202: In the case of an abnormal area in the tire tread, the abnormal area is scanned by the second detection unit to identify foreign objects in the abnormal area; wherein, the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit.

[0051] This method first uses a relatively low-frequency first detection unit to perform a preliminary inspection of the tire treads. Low-frequency ultrasound is used to determine if any abnormal areas exist within the treads. Because low-frequency ultrasound has strong penetrating power and a fast scanning speed, it can locate not only shallow but also deep abnormal areas containing foreign objects, and the localization speed is fast. This allows for a rapid and comprehensive inspection of the tire treads, and subsequent identification only requires identifying the abnormal areas. Then, a relatively high-frequency second detection unit performs a fine scan of the abnormal areas within the tire treads. High-frequency ultrasound is used to identify foreign objects within these areas, such as their shape or outline. Compared to low-frequency ultrasound, although the scanning speed of high-frequency ultrasound is relatively slower, its higher resolution and identification accuracy result in more precise identification of the shape or outline of the foreign objects. Therefore, by combining low-frequency ultrasonic scanning and high-frequency ultrasonic scanning, low-frequency ultrasonic scanning can quickly identify abnormal areas. Then, during high-frequency ultrasonic scanning, only abnormal areas can be accurately identified without scanning non-abnormal areas. This reduces the area to be scanned by high-frequency ultrasonic scanning, improves identification efficiency, and increases the detection speed and success rate of small foreign objects in tire gaps. This can reduce the safety hazards caused by foreign objects embedded in tire gaps.

[0052] In one embodiment, when there is an abnormal area in the tire tread, the abnormal area is scanned by a second detection unit to identify foreign objects in the abnormal area, including: when there is an abnormal area in the tire tread, the abnormal area is scanned by the second detection unit to obtain a target scanning signal; the target scanning signal is input into a preset foreign object identification model to identify foreign objects in the abnormal area.

[0053] In this embodiment, a foreign object recognition model, such as a stone recognition model, can be preset. When there is an abnormal area in the tire crevices, the abnormal area is scanned by the second detection unit to obtain the target scanning signal. The target scanning signal is input into the preset stone recognition model to identify foreign objects such as stones in the abnormal area.

[0054] The stone recognition model can be a lightweight convolutional neural network (CNN) architecture: input layer → time-frequency feature extraction module (3 layers) → attention mechanism → multi-scale fusion → output layer.

[0055] The input layer is a time-frequency representation of the input echo signal (e.g., 128×64 pixels).

[0056] Time-frequency feature extraction module (3 layers): Convolutional layer (extracts local features) → Activation function (such as ReLU, introducing nonlinearity) → Pooling layer (sampling, reducing parameters).

[0057] Attention mechanism: The model automatically learns to assign weights to different feature channels or image regions.

[0058] Multi-scale fusion: Feature maps are extracted at different depths of the CNN (such as after the second and third layers). Shallow feature maps have high resolution, which is beneficial for locating small targets; deep feature maps have strong semantic information, which is beneficial for recognizing large targets.

[0059] Output layer: Outputs the probability of the stone's existence (0-1) and its location estimate.

[0060] Specifically, in this embodiment, after the first detection unit completes the preliminary detection of the tire groove gap and obtains the abnormal area, the system immediately starts the secondary refined detection process, calls the second detection unit to scan the abnormal area in a targeted manner, and relies on the relatively high-frequency ultrasonic sensor of the second detection unit to perform high-precision sensing and acquisition, and completes the accurate identification and judgment of the type, size and distribution of foreign objects in the abnormal area. This makes up for the shortcomings of limited accuracy and inability to subdivide the type of foreign objects in the first round of large-scale preliminary detection, and realizes foreign object detection from regional warning to foreign object qualitative analysis.

[0061] In this embodiment, the second detection unit can be an ultrasonic detection sensor vertically positioned above the tire surface with a working frequency of 1MHz. Compared to the 400-500kHz circular array sensor in the first detection unit, the ultrasonic detection sensor in the second detection unit has advantages such as shorter wavelength, higher spatial resolution, concentrated beam directivity, and excellent short-range detection accuracy, making it suitable for applications requiring small-scale, detailed exploration of abnormal areas. The initial preliminary detection performed by the first detection unit can only detect abnormal areas, indicating the presence of local anomalies, but cannot distinguish whether the groove contains hard stones, nails, or specific foreign objects such as attached mud or condensed water droplets. However, the second detection unit, with its high-frequency detection advantage, can capture subtle differences in reflection, achieving accurate identification of foreign objects in abnormal areas.

[0062] In this embodiment, the second detection unit initiates a scanning detection of the defined abnormal area to acquire a complete target scanning signal. For example, the second detection unit directionally emits ultrasonic detection waves at an angle perpendicular to the tire surface. The sound waves penetrate directly into the abnormal area of ​​the groove, contacting the groove sidewalls, the groove bottom base, and the surfaces of various foreign objects hidden in the gaps. Because objects of different materials and shapes have significant differences in the reflection coefficient, resonance feedback characteristics, and sound wave loss of ultrasonic waves, the sound waves will generate differentiated reflected echoes and structural resonance signals after contacting the target object. The sensor then completely captures the reflected sound wave data, which is then summarized and integrated to form the target scanning signal of the abnormal area.

[0063] The acquired target scanning signal contains multiple layers of information, preserving not only the acoustic feedback of the basic structure at the anomaly location but also implicit characteristics such as the material, size, burial depth, and adhesion tightness of the foreign object. Simultaneously, the signal may contain a small amount of noise interference. Preprocessing can be performed on the acquired target scanning signal to remove clutter interference and retain the pure and effective feedback information of the foreign object and structure. Unlike the first detection unit, which focuses on filtering the overall trench waveform, the preprocessing of the target scanning signal for the second detection unit can focus on extracting parameters such as the unique resonance characteristics, subtle amplitude fluctuations, and short-term spectral changes of the high-frequency signal, further amplifying the unique acoustic characteristics of the foreign object and ensuring the data quality for subsequent model recognition.

[0064] After signal preprocessing, the target scanning signal can be formatted according to the model input specifications to obtain input data in a format that conforms to the preset standard. This data is then input into a pre-trained foreign object recognition model, which identifies foreign objects in the abnormal region. The foreign object recognition model used in this application can be a lightweight convolutional neural network (CNN) model, which has advantages such as strong feature extraction capabilities, high computational efficiency, and accurate multi-class differentiation, enabling efficient parsing of various hidden feature information behind the target scanning signal.

[0065] The overall model architecture consists of an input layer, a three-layer time-frequency feature extraction module, an attention mechanism module, a multi-scale fusion module, and a final output layer. Each layer works together to analyze signal features layer by layer. The normalized target scanning signal is first fed into the model input layer and uniformly transformed into a 128×64 time-frequency feature image. This transforms the one-dimensional scanning signal into a visualized two-dimensional feature map, intuitively presenting the signal energy distribution patterns at different times and frequencies, facilitating feature mining in subsequent layers.

[0066] After entering the three-layer time-frequency feature extraction module, the convolutional layer is used to capture local detailed features. The convolutional kernel specifically captures unique identifiers such as the edge contour of foreign objects, acoustic energy abrupt change points, and short-term waveform differences, accurately distinguishing the feature differences between the hard material of stones, the loose texture of mud, and the liquid form of water droplets. After feature extraction, a non-linear operation logic is introduced through the ReLU activation function to break the limitations of linear features, enrich the feature expression dimensions, and enable the foreign object recognition model to recognize more complex and varied foreign object morphological features. Subsequently, the pooling layer performs dimensionality reduction sampling on the feature data. On the one hand, this effectively expands the model's receptive field, taking into account the feature capture of small-volume gravel and large-area mud. On the other hand, it simplifies the model's operation parameters, reduces the proportion of redundant data, improves the overall recognition speed, and adapts to the response requirements of real-time vehicle detection.

[0067] The multi-layer feature data, processed by the three-layer feature extraction module, is then transmitted to the attention mechanism module. This module has the ability to autonomously learn weight allocation, combining the recognition experience accumulated from massive training samples to automatically determine the effectiveness of feature regions. It assigns higher weights to high-frequency, short-term, and high-energy feature channels and key image regions that can clearly distinguish the type of foreign object, while weakening redundant background features and weak noise features that have no reference value. For example, the reflection signal of hard stones is concentrated and the instantaneous energy peak is prominent, so the corresponding feature regions will be focused on by the model; while the reflection signals of mud and water droplets are dispersed and flat, and the feature weights are adjusted accordingly to strengthen the core recognition basis and reduce the interference of irrelevant factors on the judgment result.

[0068] The optimized feature data from the attention mechanism is simultaneously fed into the multi-scale fusion stage. The model extracts feature maps of different depths from the second and third layer feature extraction nodes. The shallow feature map has a significant resolution advantage, enabling precise location of small anomalies such as tiny granular stones and minor damage points within the trench. The deep feature map contains rich semantic judgment information, accurately identifying larger foreign objects such as large-scale mud accumulations and continuous water stains. The model integrates and complements the features from the two different scales, ensuring both the accuracy of foreign object location and the reliability of category determination, avoiding the missed and misclassification problems that are prone to occur in single-scale feature recognition.

[0069] After all features are analyzed and fused throughout the entire process, they are input into the model's output layer. The output layer can comprehensively compare all extracted features with standard feature sample libraries of various foreign objects, and finally output two core identification results. The first is the probability of the existence of various foreign objects in the abnormal area. The probability value ranges from 0 to 1. The closer the value is to 1, the higher the credibility of the actual existence of this type of foreign object. The second is the simultaneous output of spatial information such as the specific distribution location and approximate size range of the foreign object inside the trench and gap.

[0070] After completing the foreign object identification and judgment, the system can automatically generate a complete abnormal area detection report, clearly marking the coordinates of the abnormal point, the specific type of the identified foreign object, and distinguishing different categories such as hard stones, attached mud, and surface water droplets. It can also include information such as the size of the foreign object, its distribution range, and the hazard level assessment.

[0071] In addition, based on the identification results, targeted follow-up prompts can be made. If high-risk foreign objects such as sharp stones or nails that can easily scratch the tire tread and cause tire blowout are detected, a warning will be issued in time to remind personnel to clean them up. If they are temporary attachments such as mud or water droplets, their status will be recorded and continuously monitored. If structural damage such as rubber breakage is identified, the vehicle fault indication module can be linked to remind that maintenance is required.

[0072] In this embodiment, an abnormal area is located using a first detection unit, and a second detection unit performs fixed-point scanning to collect target scanning signals. Then, foreign objects in the abnormal area are identified according to a preset foreign object identification model. The first detection unit uses a relatively low-frequency ultrasonic sensor to quickly narrow down the anomaly detection range, while the second detection unit uses a relatively high-frequency ultrasonic sensor to identify foreign objects. This effectively solves the problem of distinguishing different debris within tire grooves. Compared to using a single-frequency ultrasonic sensor, this approach ensures both high precision and scanning speed, effectively improving the comprehensiveness and accuracy of tire hazard detection and ensuring tire safety during vehicle operation.

[0073] After the first detection unit initially determines that there is an abnormal area in the tire groove gap, in order to avoid problems such as positioning deviation, scanning omission, and inaccurate boundary determination from the root, and to effectively ensure the accuracy and completeness of the area of ​​the high-frequency scanning operation of the second detection unit, this application can perform an area expansion processing operation on the original abnormal area, and define the new area after the range expansion as the abnormal area scanned by the second detection unit. This ensures that the abnormal target and its surrounding related areas can be fully covered and detected, avoiding situations such as missed detection of foreign objects and incomplete feature extraction caused by the area being defined too narrowly.

[0074] In this embodiment, after determining the existence of an original abnormal region, the first detection unit determines the basic coordinate parameters of the original abnormal region, which serve as the basis for region expansion. For example, based on the angle data recorded during tire circumferential array detection and the tire tread plane coordinate system, the core center point position, region boundary contour, angular span range, and radial coverage size of the original abnormal region can be determined. In the tire detection plane coordinate system, with the tire tread center as the reference origin, combined with the azimuth information synchronously stored during the acquisition by the eight sets of ultrasonic sensors of the first detection unit, the minimum range contour where the signal abnormality is initially determined is determined. This contour is the original abnormal region before region expansion. The original abnormal region often only contains the most obvious core point of the abnormality. The actual coverage area of ​​the foreign object and the extension boundary of the defect may exceed the range defined by the original abnormal region. If the second detection unit directly scans according to the original abnormal region, it is very easy to encounter the problem that boundary foreign objects and extended damage cannot be captured. Therefore, the area of ​​the original abnormal region can be expanded.

[0075] In this embodiment, the area expansion processing can be carried out on the basis of the original abnormal area. For example, the outline of the original abnormal area can be expanded outward to form a new abnormal area with a slightly larger area than the original abnormal area. The new abnormal area can completely encompass the adjacent grooves and tread joints around the abnormal point, effectively capturing foreign objects such as stones, nails, and mud that are stuck on the edge of the abnormal area and closely distributed along the boundary. At the same time, it covers minor abnormal points that were not included in the initial abnormal area due to signal delay.

[0076] After determining the abnormal area after the area expansion process, the second detection unit scans according to the range corresponding to the expanded new abnormal area to obtain the target scanning signal, and inputs the target scanning signal into the preset foreign object recognition model to identify foreign objects in the abnormal area.

[0077] The anomalous area obtained through area enlargement avoids the shortcomings of the initial anomalous area, such as blurred boundaries, insufficient coverage, and easy omission of edge targets. This ensures that the second detection unit does not miss any anomalous-related points, comprehensively collecting complete acoustic feature data of foreign objects, providing a full range of dimensions and complete information for the subsequent foreign object identification model. It also effectively offsets detection positioning errors, signal judgment deviations, and positional shifts caused by tire micro-movements. From the perspective of area delineation, this improves the reliability of the target scanning signals collected by the second detection unit, reducing identification errors and missed hazard detections caused by missing scanning ranges. This ensures the accuracy of subsequent foreign object type determination, size calculation, and hazard assessment results, improving the effectiveness and comprehensiveness of foreign object detection.

[0078] In one embodiment, after identifying foreign objects in an abnormal area, the method further includes: obtaining the identification confidence level of the foreign object; marking the abnormal area when the identification confidence level is less than a first threshold so that the abnormal area can be re-scanned during the next scan; outputting a first prompt message when the identification confidence level is greater than or equal to the first threshold and less than a second threshold; wherein the first prompt message is used to indicate the presence of foreign objects in the tire gap; and outputting a second prompt message when the identification confidence level is greater than or equal to the second threshold; wherein the second prompt message is used to indicate the presence of foreign objects in the abnormal area.

[0079] In this embodiment, the identification confidence level of foreign objects can be obtained. For example, if the first threshold is 70% and the second threshold is 90%, then when the identification confidence level is <70%, the abnormal area is marked as "suspected" and will be scanned more heavily the next time the scan is started. When 70% ≤ identification confidence level <90%, a low-risk warning is issued, such as outputting a first prompt message, such as "There may be a stone in the right tire". When the identification confidence level is ≥90%, a high-risk warning is output, such as "There is a 4mm stone in the third groove of the right front wheel. It is recommended to clean it".

[0080] Based on the methods described in the above embodiments of this application, it has been verified that embedded stone particles with a size ≥3mm can be reliably detected with an accuracy rate >85%. Furthermore, by pre-setting a tire tread groove feature model, the anti-interference capability can be improved, effectively distinguishing stones from mud, water droplets, etc., reducing the false alarm rate to below 8%. This method for identifying foreign objects in tire treads can automatically complete the detection when the vehicle starts or is traveling at low speeds (<5km / h) without manual intervention. It can provide early warning before stones cause internal tire damage, preventing approximately 25% of tire failures caused by embedded foreign objects.

[0081] Based on the same technical concept, the second embodiment of this application provides a foreign object identification device in tire gaps, such as... Figure 3 The device includes: The area identification module 301 is used to perform preliminary detection of tire gaps through the first detection unit to determine whether there are abnormal areas; The foreign object identification module 302 is used to identify foreign objects in the abnormal area by scanning the abnormal area through the second detection unit when there is an abnormal area in the tire gap; wherein the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit.

[0082] The device first performs a preliminary inspection of the tire treads using a relatively low-frequency first detection unit. Low-frequency ultrasound is used to determine if any abnormal areas exist within the treads. Because low-frequency ultrasound has strong penetrating power and a fast scanning speed, it can locate both shallow and deep abnormal areas containing foreign objects, and the localization speed is fast. This allows for a rapid and comprehensive inspection of the tire treads, and subsequent identification only requires identifying the abnormal areas. Then, a relatively high-frequency second detection unit performs a fine scan of the abnormal areas within the tire treads. High-frequency ultrasound is used to identify foreign objects within these areas, such as their shape or outline. Compared to low-frequency ultrasound, although the scanning speed of high-frequency ultrasound is relatively slower, its higher resolution and identification accuracy result in more precise identification of the shape or outline of the foreign objects. Therefore, by combining low-frequency ultrasonic scanning and high-frequency ultrasonic scanning, low-frequency ultrasonic scanning can quickly identify abnormal areas. Then, during high-frequency ultrasonic scanning, only abnormal areas can be accurately identified without scanning non-abnormal areas. This reduces the area to be scanned by high-frequency ultrasonic scanning, improves identification efficiency, and increases the detection speed and success rate of small foreign objects in tire gaps. This can reduce the safety hazards caused by foreign objects embedded in tire gaps.

[0083] Optionally, the area recognition module 301 is specifically used to continuously monitor the current vehicle speed; when the current vehicle speed is less than a preset threshold, the first detection unit performs a preliminary detection of the tire gap to determine whether there is an abnormal area.

[0084] Optionally, the first detection unit includes multiple first sensors disposed on the wheel arch, and a region identification module 301, specifically used to collect echo signals from the tire tread groove once at preset rotational step intervals using the first sensors; to perform beamforming on all echo signals to obtain an image inside the groove; and to determine whether there are abnormal regions in the image inside the groove. Optionally, to perform beamforming on all echo signals to obtain an image inside the groove includes: aligning the target echo signals collected by each first sensor to obtain an aligned signal; and performing beamforming on the aligned signals of all first sensors to obtain an image inside the groove. Optionally, before aligning the target echo signals collected by each first sensor, the region identification module 301 is further used to: acquire a preset tire tread groove feature model; filter the original echo signals collected by the first sensors according to the tire tread groove feature model to obtain a clean echo signal; and enhance the foreign object features in the clean echo signal to obtain a processed target echo signal. Optionally, feature enhancement is performed on the foreign object features in the pure echo signal to obtain the processed target echo signal, including: obtaining a preset enhancement coefficient; performing feature enhancement on the foreign object features based on the tire tread groove feature model, the enhancement coefficient, and the pure echo signal to obtain the processed target echo signal.

[0085] Optionally, the foreign object recognition module 302 is specifically used to scan the abnormal area through the second detection unit to obtain the target scanning signal when there is an abnormal area in the tire gap; and to input the target scanning signal into a preset foreign object recognition model to identify the foreign object in the abnormal area.

[0086] Optionally, after identifying foreign objects in the abnormal area, the foreign object identification module 302 is further configured to: obtain the identification confidence level of the foreign object; mark the abnormal area when the identification confidence level is less than a first threshold so that the abnormal area can be re-scanned during the next scan; output a first prompt message when the identification confidence level is greater than or equal to the first threshold and less than a second threshold; wherein the first prompt message is used to indicate that there is a foreign object in the tire gap; and output a second prompt message when the identification confidence level is greater than or equal to the second threshold; wherein the second prompt message is used to indicate that there is a foreign object in the abnormal area.

[0087] like Figure 4 As shown in the figure, this application embodiment provides a vehicle, including a processor 111, a communication interface 112, a memory 113, and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114. Memory 113 is used to store computer programs; In one embodiment of this application, the processor 111, when executing the program stored in the memory 113, implements the foreign object identification method in tire gaps provided in any of the aforementioned method embodiments.

[0088] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0089] The communication interface is used for communication between the aforementioned terminal and other devices.

[0090] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0091] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0092] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for identifying foreign objects in tire gaps as provided in any of the foregoing method embodiments.

[0093] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0094] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0095] It should be understood that the terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “described” as used herein may also include the plural forms. The terms “comprising,” “including,” “containing,” and “having” are inclusive and therefore indicate the presence of the stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not construed as requiring them to be performed in a particular order described or illustrated unless the order of performance is explicitly indicated. It should also be understood that additional or alternative steps may be used.

[0096] It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. In the description, suffixes such as "module," "part," or "unit" used to denote elements are used solely for illustrative purposes and have no specific meaning in themselves. Therefore, "module," "part," or "unit" may be used interchangeably.

[0097] The above description is merely a specific embodiment of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for identifying foreign objects in tire gaps, characterized in that, The method includes: The first detection unit performs a preliminary inspection of the tire gaps to determine if any abnormal areas exist. In the event of an abnormal area in the tire tread, a second detection unit scans the abnormal area to identify foreign objects within it; wherein, the second scanning frequency of the second detection unit is greater than the first scanning frequency of the first detection unit.

2. The method according to claim 1, characterized in that, The first detection unit performs a preliminary inspection of the tire gaps to determine if any abnormal areas exist, including: Continuously monitor the current vehicle speed; When the current vehicle speed is less than a preset threshold, the first detection unit performs a preliminary detection of the tire gaps to determine whether there are any abnormal areas.

3. The method according to claim 1, characterized in that, The first detection unit includes multiple first sensors disposed on the wheel arch; the first detection unit performs preliminary detection of tire gaps to determine whether there are abnormal areas, including: The first sensor collects an echo signal from the tire gap once at preset rotation intervals; All echo signals are beamformed to obtain an image of the inside of the trench; Determine whether there are abnormal areas in the image inside the trench.

4. The method according to claim 3, characterized in that, All echo signals are beamformed to obtain an image of the trench interior, including: Each target echo signal collected by the first sensor is aligned to obtain an alignment signal; The alignment signals from all the first sensors are beamformed to obtain an image of the inside of the trench.

5. The method according to claim 4, characterized in that, Before aligning the target echo signals acquired by each of the first sensors, the method further includes: Obtain the preset tire tread groove feature model; The original echo signal collected by the first sensor is filtered according to the tire tread groove feature model to obtain a clean echo signal; The foreign object features in the pure echo signal are enhanced to obtain the processed target echo signal.

6. The method according to claim 5, characterized in that, Feature enhancement is performed on the foreign object features in the pure echo signal to obtain the processed target echo signal, including: Obtain the preset enhancement coefficient; The foreign object feature is enhanced based on the tire tread groove feature model, the enhancement coefficient, and the pure echo signal to obtain the processed target echo signal.

7. The method according to claim 1, characterized in that, In the event of an abnormal area in the tire tread, a second detection unit scans the abnormal area to identify foreign objects within it, including: In the event of an abnormal area in the tire tread, the abnormal area is scanned by the second detection unit to obtain the target scanning signal; The target scanning signal is input into a preset foreign object identification model to identify foreign objects in the abnormal area.

8. The method according to claim 1, characterized in that, After identifying foreign objects in the abnormal area, the method further includes: Obtain the recognition confidence level of the foreign object; When the identification confidence level is less than a first threshold, the abnormal region is marked so that it can be re-scanned during the next scan. When the identification confidence level is greater than or equal to the first threshold and less than the second threshold, a first prompt message is output; wherein, the first prompt message is used to indicate that there is a foreign object in the tire gap; When the identification confidence level is greater than or equal to the second threshold, a second prompt message is output; wherein, the second prompt message is used to indicate that there is a foreign object in the abnormal area.

9. A vehicle, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor, when executing a program stored in the memory, implements the method for identifying foreign objects in tire gaps as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for identifying foreign objects in tire gaps as described in any one of claims 1-8.

Images