Computer vision-based vehicle dynamic weighing method for side toll station

Abstract

This invention discloses a dynamic vehicle weighing method for side-mounted toll stations based on computer vision. The method includes the following steps: Step 1: Installing a dynamic weighing device; Step 2: Vehicle identification; Step 3: Calculating the axle weight of the first axle, including: Step 3-1: Identifying the number of tires; Step 3-2: Acquiring tire sidewall images, which include an industrial tire sidewall image, an infrared tire sidewall image, and a fused tire sidewall image; the fused tire sidewall image is a fusion of the industrial tire sidewall image and the infrared tire sidewall image; Step 3-3: Calculating the axle weight of the first axle; Step 3-4: Calculating the total axle weight of the first axle; Step 4: Calculating the axle weight of the i-th axle; Step 5: Calculating the total weight of the vehicle to be weighed. This invention enables accurate dynamic weighing of vehicles without causing excavation damage to the toll station road surface. The weighing is unaffected by time and weather, and has low maintenance costs, high efficiency, high accuracy, easy information integration, and strong stability.

Description

Technical Field

[0001] This invention relates to the field of traffic engineering, and in particular to a dynamic vehicle weighing method for side-mounted toll stations based on computer vision. Background Technology

[0002] Statistics show that my country's comprehensive transportation network has exceeded 6 million kilometers in length, with expressways covering over 98% of cities with a population of over 200,000. However, with the rapid development of my country's transportation infrastructure, various overloading problems have also emerged. According to the Ministry of Transport, highways, bridges, and other infrastructure in my country are currently operating under conditions of overloading and exceeding weight limits. Illegal overloading has become the leading cause of road accidents, severely damaging highway and bridge facilities, easily triggering traffic accidents, and endangering people's lives and property. Furthermore, this illegal behavior seriously disrupts the transportation market order. It has been proven that over 80% of road traffic accidents involving heavy trucks are caused by overloading. Therefore, to effectively curb overloading, ensure road traffic safety, maintain a sound transportation market order, and promote the sustainable and healthy development of my country's economy, timely and accurate acquisition of vehicle tire load and vehicle load is particularly important.

[0003] Currently, the mainstream vehicle weighing technologies include static weighbridges, bridge dynamic weighing systems, and road dynamic weighing systems. However, these technologies still have the following problems in their application:

[0004] 1. Static weighbridges require the installation of weighing service stations, which are complex to construct, have high installation and operating costs, and require frequent maintenance. Furthermore, they are slow, inefficient, and cannot perform dynamic weighing, easily causing traffic congestion during peak hours.

[0005] 2. Bridge dynamic weighing systems require the installation of sensors on the bridge, resulting in relatively high equipment costs and significant capital investment. Furthermore, the measurement results are greatly affected by the sensor placement, necessitating specific sensor locations and making subsequent maintenance and replacement difficult. Additionally, this method weighs vehicles as they travel across the bridge, failing to prevent overloaded vehicles from crossing.

[0006] 3. The road dynamic weighing system requires sensors to be installed on the road surface to measure the instantaneous dynamic force when a moving vehicle passes the sensor. The static axle load is then obtained through signal processing, thus determining the vehicle's weight. This system requires traffic interruption and the excavation of concrete pits or grooves in the road surface during installation and subsequent maintenance, resulting in low efficiency. Furthermore, its measurement accuracy is significantly affected by the road surface stiffness and unevenness.

[0007] In summary, improving the speed and accuracy of tire load and vehicle load measurement, reducing the impact of environmental factors on measurement results, lowering the cost and subsequent maintenance costs, and enhancing the system's all-weather usability are technical problems that urgently need to be solved by those skilled in the art. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to address the shortcomings of the prior art by providing a computer vision-based dynamic vehicle weighing method for side-mounted toll stations. This computer vision-based dynamic vehicle weighing method for side-mounted toll stations can achieve accurate dynamic weighing of vehicles without causing excavation damage to the toll station road surface, and the weighing is not affected by time or weather.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0010] A computer vision-based method for dynamic vehicle weighing at side-mounted toll stations includes the following steps.

[0011] Step 1: Install the dynamic weighing device: The dynamic weighing device includes a piezoelectric tire recognition sensor, a vehicle recognition camera, and a tire image recognition camera system; the piezoelectric tire recognition sensor is laid laterally in the toll station's driving lane, and the laying length does not exceed half the width of the corresponding driving lane; the vehicle recognition camera is set above the toll station's driving lane; the tire image recognition camera system is mounted on one side of the toll station's driving lane and includes a flash, an industrial camera, and an infrared thermal imaging camera.

[0012] Step 2, Vehicle Recognition: When a vehicle to be weighed arrives at the toll station, the vehicle recognition camera identifies the license plate of the vehicle.

[0013] Step 3: Calculate the axle load of the first wheel axle, which includes the following steps:

[0014] Step 3-1: Identify the number of tires: When a single tire on the first axle passes the piezoelectric tire identification sensor, the piezoelectric tire identification sensor identifies whether the single tire on the first axle is a single tire or a double tire through the piezoelectric signal.

[0015] Step 3-2: Acquire tire sidewall images, including industrial tire sidewall images, infrared tire sidewall images, and fused tire sidewall images. The specific acquisition method includes the following steps:

[0016] Step 3-2a, Fill light: Use a flash to fill light on the outer side of the tire on one side of the first axle.

[0017] Step 3-2b: Take an industrial image of the tire sidewall: While the flash is flashing, the industrial camera takes a picture of the outer sidewall of a single tire on the first axle to obtain an optical image of the tire sidewall.

[0018] Step 3-2c: Capture the infrared image of the tire sidewall: While the flash is flashing, the infrared thermal imaging camera takes a picture of the outer sidewall of the tire on one side of the first axle to obtain the infrared image of the tire sidewall.

[0019] Step 3-2d: Obtain the tire sidewall fusion image: Fuse the industrial image of the tire sidewall and the infrared image of the tire sidewall to obtain the tire sidewall fusion image.

[0020] Step 3-3: Calculate the axle load for the first axle category: Using a deep learning model, the tire sidewall image obtained in Step 3-2 is used as the input to the deep learning model, and the output is the axle load for the first axle category; the axle load for the first axle category includes the industrial axle load W. 1a First wheel axle infrared axle load W 1b Axle load W integrated with the first wheel axle 1c Among them, the first wheel axle industrial axle load W 1a First wheel axle infrared axle load W 1b Axle load W integrated with the first wheel axle 1c The corresponding inputs in the deep learning model are the industrial image of the tire sidewall, the infrared image of the tire sidewall, and the fused image of the tire sidewall, respectively.

[0021] Steps 3-4: Calculate the total axle load W1 of the first wheel axle. The specific calculation formula is as follows:

[0022]

[0023] in, The weighted axle load for the first axle is calculated using the following formula:

[0024]

[0025] In the formula, m and n are W 1a and W 1b The weighting coefficients are m > n and m + n = 1.

[0026] Step 4: Calculate the axle load of the i-th wheel axle: Repeat step 3 to obtain the total axle load W of the i-th wheel axle. i Where 1 < i < j, and j is the total number of wheel axles of the vehicle to be weighed.

[0027] Step 5: Calculate the total weight W of the vehicle to be weighed.

[0028]

[0029] In step 1, the industrial camera and the infrared thermal imaging camera are positioned on both sides of the flash lamp, and are equidistant from the center of the tire to be weighed.

[0030] In step 3-3, the deep learning model is the XGboost load prediction model.

[0031] In step 3-3, when using the XGboost load prediction model to predict the axle load of the first wheel axle category, the tire side image needs to be preprocessed to obtain tire features; then the tire features are used as input to the XGboost load prediction model; the tire features include tire character recognition results and tire vertical deflection information; the tire character recognition results include tire size, tire model and tire pressure.

[0032] In step 3-3, the tire features include industrial image tire features, infrared image tire features, and fused image tire features; the industrial image tire features, infrared image tire features, and fused image tire features all include tire character recognition results and tire vertical deflection information; the industrial image tire features, infrared image tire features, and fused image tire features correspond to the tire side industrial image, tire side infrared image, and tire side fused image, respectively.

[0033] The tire character recognition results are obtained by detecting the corresponding tire side image through OCR recognition algorithm; the tire vertical deflection information is obtained by detecting geometric fitting and region growing algorithms through image iteration.

[0034] In step 3-2d, the resolution of the industrial camera is higher than that of the infrared thermal imaging camera. Therefore, the fusion method for tire sidewall images includes the following steps:

[0035] Step 3-2d1: Construct the DDcGAN neural network model: The input to the DDcGAN neural network model is an industrial image of the tire sidewall and an infrared image of the tire sidewall with the same resolution as the industrial camera; the output of the DDcGAN neural network model is a fused tire sidewall image; the fused tire sidewall image has the same resolution of texture detail as the industrial tire sidewall image and the same pixel value amplitude of temperature saliency as the infrared image of the tire sidewall.

[0036] Step 3-2d2: Improve the resolution of the tire sidewall infrared image: Use the convolutional layer trained in the DDcGAN neural network model to upsample the low-resolution tire sidewall infrared image from step 3-2c to the resolution of the industrial camera, forming a high-resolution tire sidewall infrared image.

[0037] Step 3-2d3, Image Fusion: The tire sidewall industrial image from Step 3-2b and the high-resolution tire sidewall infrared image from Step 3-2d2 are used as inputs to the DDcGAN neural network model constructed in Step 3-2d1 to obtain the fused tire sidewall image.

[0038] In steps 3-4, m = 0.8 and n = 0.2.

[0039] In step 3-2b, the industrial camera has 12 million pixels and uses a sampling frequency of 5Hz to acquire industrial images of the tire sidewall.

[0040] In step 3-2c, the infrared thermal imaging camera has a resolution of 640×512 pixels and acquires an infrared image of the tire sidewall at a sampling frequency of 5Hz.

[0041] The present invention has the following beneficial effects:

[0042] 1. This invention enables accurate dynamic weighing of vehicles without causing excavation damage to the toll station road surface.

[0043] 2. At night or in dim weather, by fusing industrial images and infrared images of the tire sidewall, a high-resolution tire sidewall fused image with significant temperature-related pixel value ranges is obtained. Tire features are extracted from the fused tire sidewall image, and then deep learning is performed through a machine learning model to obtain the corresponding axle load. Therefore, the weighing can be unaffected by time and weather.

[0044] 3. This invention has a wide range of applications, is easy to integrate information, and compared with using only an optical camera, the device using the combination of an optical camera and an infrared camera can work stably for a long time in all weather conditions. Attached Figure Description

[0045] Figure 1 The flowchart of the present invention, which is a computer vision-based method for dynamic vehicle weighing at a side-mounted toll station, is shown.

[0046] Figure 2 The diagram shows the layout of the piezoelectric tire recognition sensor in this invention.

[0047] Figure 3 A schematic diagram of the tire image recognition camera system of the present invention is shown.

[0048] The components include: 1. Piezoelectric tire recognition sensor; 2. Tire image recognition camera system; 3. Display screen; 4. Infrared thermal imaging camera; 5. Flash lamp; 6. Industrial camera; 7. Central controller; 8. Deep learning development board; 9. Power supply. Detailed Implementation

[0049] The present invention will now be described in further detail with reference to the accompanying drawings and specific preferred embodiments.

[0050] In the description of this invention, it should be understood that the terms "left side," "right side," "upper part," "lower part," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. "First," "second," etc., do not indicate the importance of the components, and therefore should not be construed as a limitation of this invention. The specific dimensions used in this embodiment are only for illustrating the technical solution and do not limit the scope of protection of this invention.

[0051] like Figure 1 As shown, a computer vision-based method for dynamic vehicle weighing at a side-mounted toll station includes the following steps.

[0052] Step 1: Install the dynamic weighing device

[0053] like Figure 2 and Figure 3 As shown, the dynamic weighing device includes a piezoelectric tire recognition sensor 1, a vehicle recognition camera, and a tire image recognition camera system 2.

[0054] Piezoelectric tire recognition sensors are laid laterally in the driving lanes of the toll station, and the length of the laying does not exceed half the width of the corresponding driving lane.

[0055] Vehicle recognition cameras are installed above the tollbooth lanes, and existing cameras at the tollbooth can be used directly, eliminating the need for additional installations and saving costs.

[0056] The tire image recognition camera system is installed on one side of the toll station's driving lane and includes a display screen 3, a flash 5, an industrial camera 6, an infrared thermal imaging camera 4, a central controller 7, a deep learning development board 8, and a power supply 9.

[0057] The aforementioned industrial camera and infrared thermal imaging camera are preferably symmetrically arranged on both sides of the flash lamp, and equidistant from the center of the tire to be weighed. The industrial camera preferably has 12 megapixels and acquires industrial images of the tire sidewall using a sampling frequency of 5Hz. The infrared thermal imaging camera preferably has 640×512 pixels and can acquire infrared images of the tire sidewall using a sampling frequency of 5Hz.

[0058] The aforementioned display screen can show information such as the weight of the vehicle to be weighed.

[0059] Step 2, Vehicle Recognition: When a vehicle to be weighed arrives at the toll station, the vehicle recognition camera identifies the license plate of the vehicle.

[0060] Step 3: Calculate the axle load of the first wheel axle, which includes the following steps:

[0061] Step 3-1: Identify the number of tires: When a single tire on the first axle passes the piezoelectric tire identification sensor, the piezoelectric tire identification sensor identifies whether the single tire on the first axle is a single tire or a double tire through the piezoelectric signal.

[0062] Step 3-2: Acquire tire sidewall images, including industrial tire sidewall images, infrared tire sidewall images, and fused tire sidewall images. The specific acquisition method includes the following steps:

[0063] Step 3-2a, Fill light: Use a flash to fill light on the outer side of the tire on one side of the first axle.

[0064] Step 3-2b: Take an industrial image of the tire sidewall: While the flash is flashing, the industrial camera takes a picture of the outer sidewall of a single tire on the first axle to obtain an optical image of the tire sidewall.

[0065] Step 3-2c: Capture the infrared image of the tire sidewall: While the flash is flashing, the infrared thermal imaging camera takes a picture of the outer sidewall of the tire on one side of the first axle to obtain the infrared image of the tire sidewall.

[0066] Step 3-2d: Obtain the tire sidewall fusion image: Fuse the industrial image of the tire sidewall and the infrared image of the tire sidewall to obtain the tire sidewall fusion image.

[0067] Since the resolution of industrial cameras is higher than that of infrared thermal imaging cameras, the above-mentioned tire sidewall image fusion method includes the following steps:

[0068] Step 3-2d1: Construct the DDcGAN neural network model: The input to the DDcGAN neural network model is an industrial image of the tire sidewall and an infrared image of the tire sidewall with the same resolution as the industrial camera; the output of the DDcGAN neural network model is a fused tire sidewall image; the fused tire sidewall image has the same resolution of texture detail as the industrial tire sidewall image and the same pixel value amplitude of temperature saliency as the infrared image of the tire sidewall.

[0069] Step 3-2d2: Improve the resolution of the tire sidewall infrared image: Use the convolutional layer trained in the DDcGAN neural network model to upsample the low-resolution tire sidewall infrared image from step 3-2c to the resolution of the industrial camera, forming a high-resolution tire sidewall infrared image.

[0070] Step 3-2d3, Image Fusion: The tire sidewall industrial image from Step 3-2b and the high-resolution tire sidewall infrared image from Step 3-2d2 are used as inputs to the DDcGAN neural network model constructed in Step 3-2d1 to obtain the fused tire sidewall image.

[0071] Step 3-3: Calculate the axle load for the first axle category: Using a deep learning model, the tire sidewall image obtained in Step 3-2 is used as the input to the deep learning model, and the output is the axle load for the first axle category; the axle load for the first axle category includes the industrial axle load W. 1a First wheel axle infrared axle load W 1b Axle load W integrated with the first wheel axle 1c Among them, the first wheel axle industrial axle load W 1a First wheel axle infrared axle load W 1b Axle load W integrated with the first wheel axle 1c The corresponding inputs in the deep learning model are the industrial image of the tire sidewall, the infrared image of the tire sidewall, and the fused image of the tire sidewall, respectively.

[0072] The aforementioned deep learning model is preferably the XGboost load prediction model, which is built into the deep learning development board.

[0073] When using the XGboost load prediction model to predict the axle load of the first wheel axle category, the tire sidewall image needs to be preprocessed to obtain tire features; then the tire features are used as input to the XGboost load prediction model; the tire features include tire character recognition results and tire vertical deflection information; the tire character recognition results include tire size, tire model and tire pressure.

[0074] Furthermore, the tire features include industrial image tire features, infrared image tire features, and fused image tire features; the industrial image tire features, infrared image tire features, and fused image tire features all include tire character recognition results and tire vertical deflection information; the industrial image tire features, infrared image tire features, and fused image tire features correspond to the tire sidewall industrial image, tire sidewall infrared image, and tire sidewall fused image, respectively.

[0075] The tire character recognition results are preferably obtained by detecting the corresponding tire side image using an OCR recognition algorithm; the tire vertical deflection information is preferably obtained by detecting geometric fitting and region growing algorithms through image iteration.

[0076] Steps 3-4: Calculate the total axle load W1 of the first wheel axle. The specific calculation formula is as follows:

[0077]

[0078] in, The weighted axle load for the first axle is calculated using the following formula:

[0079]

[0080] In the formula, m and n are W 1a and W 1b The weighting coefficients are m > n and m + n = 1. Further optimization is performed with m = 0.8 and n = 0.2.

[0081] Step 4: Calculate the axle load of the i-th wheel axle: Repeat step 3 to obtain the total axle load W of the i-th wheel axle. i Where 1 < i < j, and j is the total number of wheel axles of the vehicle to be weighed.

[0082] Step 5: Calculate the total weight W of the vehicle to be weighed.

[0083]

[0084] Finally, the tire calculation results and W are displayed together on the screen.

[0085] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A computer vision-based method for dynamic vehicle weighing at a side-mounted toll station, characterized in that: Includes the following steps: Step 1: Install the dynamic weighing device: The dynamic weighing device includes a piezoelectric tire recognition sensor, a vehicle recognition camera, and a tire image recognition camera system; the piezoelectric tire recognition sensor is laid laterally in the toll station's driving lane, and the laying length does not exceed half the width of the corresponding driving lane; the vehicle recognition camera is set above the toll station's driving lane; the tire image recognition camera system is mounted on one side of the toll station's driving lane and includes a flash, an industrial camera, and an infrared thermal imaging camera; Step 2, Vehicle Recognition: When a vehicle to be weighed arrives at the toll station, the vehicle recognition camera identifies the license plate of the vehicle to be weighed. Step 3: Calculate the axle load of the first wheel axle, which includes the following steps: Step 3-1, Identify the number of tires: When a single tire on one side of the first axle passes the piezoelectric tire identification sensor, the piezoelectric tire identification sensor identifies whether the single tire on the first axle is a single tire or a double tire through the piezoelectric signal. Step 3-2: Acquire tire sidewall images, including industrial tire sidewall images, infrared tire sidewall images, and fused tire sidewall images. The specific acquisition method includes the following steps: Step 3-2a, Fill light: Use a flash to fill light on the outer side of the tire on one side of the first axle; Step 3-2b: Take an industrial image of the tire sidewall: While the flash is flashing, the industrial camera takes a picture of the outer sidewall of a single tire on the first axle to obtain an optical image of the tire sidewall. Step 3-2c: Take an infrared image of the tire sidewall: While the flash is flashing, the infrared thermal imaging camera takes a picture of the outer sidewall of one tire on the first axle to obtain an infrared image of the tire sidewall. Step 3-2d: Obtain the tire sidewall fusion image: Fuse the industrial image of the tire sidewall and the infrared image of the tire sidewall to obtain the tire sidewall fusion image; Step 3-3: Calculate the axle load for the first axle category: Using a deep learning model, the tire sidewall image obtained in Step 3-2 is used as the input to the deep learning model, and the output is the axle load for the first axle category; the axle load for the first axle category includes the industrial axle load W. 1a First wheel axle infrared axle load W 1b Axle load W integrated with the first wheel axle 1c Among them, the first wheel axle industrial axle load W 1a First wheel axle infrared axle load W 1b Axle load W integrated with the first wheel axle 1c The corresponding inputs in the deep learning model are industrial images of tire sidewalls, infrared images of tire sidewalls, and fused images of tire sidewalls, respectively. Steps 3-4: Calculate the total axle load W1 of the first wheel axle. The specific calculation formula is as follows: in, The weighted axle load for the first axle is calculated using the following formula: In the formula, m and n are W 1a and W 1b The weighting coefficients, m > n and m + n = 1; Step 4: Calculate the axle load of the i-th wheel axle: Repeat step 3 to obtain the total axle load W of the i-th wheel axle. i Where 1 < i < j, and j is the total number of wheel axles of the vehicle to be weighed; Step 5: Calculate the total weight W of the vehicle to be weighed.

2. The computer vision-based vehicle dynamic weighing method for side-mounted toll stations according to claim 1, characterized in that: In step 1, the industrial camera and the infrared thermal imaging camera are positioned on both sides of the flash lamp, and are equidistant from the center of the tire to be weighed.

3. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 1, characterized in that: In step 3-3, the deep learning model is the XGboost load prediction model.

4. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 3, characterized in that: In step 3-3, when using the XGboost load prediction model to predict the axle load of the first wheel axle category, it is necessary to first identify the tire side image to obtain tire features; use the tire features as input to the XGboost load prediction model; the tire features include tire character recognition results and tire vertical deflection information; the tire character recognition results include tire size, tire model and tire pressure.

5. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 4, characterized in that: In step 3-3, the tire features include industrial image tire features, infrared image tire features, and fused image tire features; the industrial image tire features, infrared image tire features, and fused image tire features all include tire character recognition results and tire vertical deflection information; the industrial image tire features, infrared image tire features, and fused image tire features correspond to the tire side industrial image, tire side infrared image, and tire side fused image, respectively.

6. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 4, characterized in that: The tire character recognition results are obtained by detecting the corresponding tire side image through OCR recognition algorithm; the tire vertical deflection information is obtained by detecting geometric fitting and region growing algorithms through image iteration.

7. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 1, characterized in that: In step 3-2d, the resolution of the industrial camera is higher than that of the infrared thermal imaging camera. Therefore, the fusion method for tire sidewall images includes the following steps: Step 3-2d1: Construct the DDcGAN neural network model: The input to the DDcGAN neural network model is an industrial image of the tire sidewall and an infrared image of the tire sidewall with the same resolution as the industrial camera; the output of the DDcGAN neural network model is a fused image of the tire sidewall; the fused image of the tire sidewall has the same resolution of texture details as the industrial image of the tire sidewall and the same pixel value amplitude of temperature saliency as the infrared image of the tire sidewall. Step 3-2d2: Improve the resolution of the tire sidewall infrared image: Use the convolutional layer trained in the DDcGAN neural network model to upsample the low-resolution tire sidewall infrared image in step 3-2c to the resolution of the industrial camera, forming a high-resolution tire sidewall infrared image. Step 3-2d3, Image Fusion: The tire sidewall industrial image from Step 3-2b and the high-resolution tire sidewall infrared image from Step 3-2d2 are used as inputs to the DDcGAN neural network model constructed in Step 3-2d1 to obtain the fused tire sidewall image.

8. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 1, characterized in that: In steps 3-4, m = 0.8 and n = 0.

2.

9. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 1, characterized in that: In step 3-2b, the industrial camera has 12 million pixels and uses a sampling frequency of 5Hz to acquire industrial images of the tire sidewall.

10. The computer vision-based dynamic vehicle weighing method for side-mounted toll stations according to claim 1, characterized in that: In step 3-2c, the infrared thermal imaging camera has a resolution of 640×512 pixels and acquires an infrared image of the tire sidewall at a sampling frequency of 5Hz.

Images