Artificial Intelligence Supervision Methods for Exhaust Gas Detection

Abstract

The artificial intelligence supervision method for exhaust gas detection is to monitor the display screen of motor vehicle exhaust gas detection equipment through a camera, use artificial intelligence technology to identify the detection data and information in the video stream collected by the camera, take the identified display value as the real value, and transmit it to the supervision department for online supervision through the Internet of Things technology. During the detection process, dynamic instantaneous images in the video stream are obtained at a specified sampling frequency, the information and display data in the dynamic instantaneous images are identified, and the recognition detection data of each dynamic instantaneous image obtained or the recognition detection data every 1 second or the recognition detection result data within the specified data acquisition time are transmitted to facilitate comparison with the detection process data or detection report result data corresponding to the detection control system. It is suitable for artificial intelligence automatic supervision of exhaust gas detection in use motor vehicles or daily calibration of exhaust gas detection equipment.

Description

Technical Field

[0001] The artificial intelligence supervision method for exhaust gas detection is a method that monitors the display screen of motor vehicle exhaust gas detection equipment through a camera, uses artificial intelligence technology to identify the detection data and information in the video stream collected by the camera, and transmits it to the regulatory department for online supervision through the Internet of Things technology. It belongs to the technical field of artificial intelligence for automobile performance testing. Background Art

[0002] In the regular inspection of exhaust emissions from in-use motor vehicles, in order to ensure the authenticity, objectivity, and accuracy of the inspection, the existing technical supervision method is: 1. The inspection station transmits the inspection process data and the inspection result data to the supervision department, and the result data is supervised by the process data. Due to various reasons, the uploaded inspection data may not match the actual inspection data, making it difficult to ensure the authenticity of the inspection result data; 2. The inspection equipment is monitored through video, with the camera facing the display screen interface of the inspection equipment to facilitate manual monitoring and collection of the operation status of the inspection equipment during the process. This manual supervision is only suitable for spot checks and is affected by subjective factors. It also has high labor costs and low efficiency. Since the exhaust gas detection equipment undergoes regular inspection and daily calibration, its display screen data has good authenticity and accuracy. Therefore, in order to overcome the shortcomings of the existing technology, the use of exhaust gas detection artificial intelligence supervision methods can greatly improve the comprehensiveness and efficiency of supervision. Summary of the Invention

[0003] The AI-powered supervision method for exhaust gas detection uses cameras to capture the video stream from the display screen of the exhaust gas detection equipment at the testing station. AI image recognition technology is then used to identify the detection information and data contained in the images in the video stream and transmit it to the supervisory department. This information is then automatically compared with the test process data or reported test result data transmitted by the testing station to the supervisory department. When the comparison data differ significantly, an alarm is triggered, thus achieving AI-powered supervision. The detection data captured by the AI-powered camera in the video stream is referred to as "identified detection data." The national standard stipulates that the identification detection result data within the acquisition time is referred to as "identified detection result data." The supervisory department can be an internal department within the testing station or an administrative department.

[0004] For video stream image acquisition, dynamic instantaneous images can be captured from the start to the end of the test, from the sampling start and end signals emitted by the test control system, or from the time the test equipment displays the sampling process. The standard stipulates that the recording period for periodic vehicle exhaust emission testing process data is generally one second, and the sampling frequency of dynamic instantaneous video stream images is set within a range of 1 to 60. Artificial intelligence image recognition technology is used to identify the test-related information and data within each image. The recognition test data for that second is the single recognition test data, the average of several recognition test data within each second, or the maximum recognition test data within each second. The recognition test data for each dynamic instantaneous image, the recognition test data for each second, or the recognition test result data within the standard's specified acquisition time can be transmitted, and the corresponding time of the recognition test result data is recorded to facilitate automatic comparison with the corresponding test process data or test report result data from the test control system, implementing artificial intelligence automated supervision.

[0005] The AI-powered monitoring method for exhaust gas detection is a method for AI-powered automatic monitoring of motor vehicle exhaust emission detection. It features: a video stream from the display screen of the motor vehicle exhaust gas detection equipment is captured via a camera, and AI image recognition technology is used to identify detection information and data in the captured image. The identified display screen values ​​are used as true values ​​and transmitted to the monitoring department. Dynamic instantaneous images are acquired from the start to the end of the detection process, or during the period of two sampling start and end signals issued by the detection control system, or during the period of the detection process displayed by the detection equipment. The sampling frequency of the dynamic instantaneous images is in the range of 1 to 60, and relevant detection information and data in the captured images are identified. The identification detection data for each dynamic instantaneous image, the average value of the identification detection data within each second, or the maximum value of the identification detection data within each second is used as the identification detection data for that second. The identification detection data for each dynamic instantaneous image, the identification detection data for each second, or the identification detection result data within the data acquisition time specified by the standard are transmitted for comparison with the detection process data or test report result data corresponding to the detection control system. The method is suitable for AI-powered automatic monitoring of in-use motor vehicle exhaust gas detection or routine calibration of exhaust gas detection equipment.

[0006] Exhaust gas detection equipment display data that has undergone verification and routine calibration is generally authentic, objective, and accurate. Using AI-powered monitoring of the detection process or test report results using display data can significantly improve the comprehensiveness and effectiveness of monitoring. This technology not only enables AI-powered monitoring of regular exhaust emissions testing for in-use motor vehicles, but also enables comparison of exhaust gas detection equipment display data with reference material data during routine calibration, ensuring the standardization of routine AI-powered monitoring of exhaust gas detection equipment. DETAILED DESCRIPTION

[0007] Using the detection method specified in the exhaust emission detection standard for in-use motor vehicles, a camera is used to monitor the display screen of the detection equipment. The dynamic instantaneous image captured by the camera during the sampling start and end period is obtained according to the two signals sent by the detection control system. Artificial intelligence image recognition technology is used to identify the detection information and data in the collected image, and the recognition detection result data within the sampling time specified in the standard is transmitted to illustrate the specific implementation method of the exhaust gas detection artificial intelligence supervision method.

[0008] 1. The steady-state operating method, dual-idle method, and loaded deceleration method for gasoline vehicles use the average value of the detection data within the sampling time specified in the standard as the detection result data. The sampling frequency is 5, and the average value of the identification detection data within each second is used as the identification detection data for that second. The average value of the identification detection data within the sampling time specified in the standard is used as the identification detection result data and transmitted.

[0009] 2. The simple transient operating condition method for gasoline vehicles uses the cumulative value of the detection data within the standard sampling time as the detection result data. The sampling frequency is 5, and the average value of the identification detection data within each second is used as the identification detection data for that second. The cumulative value of the identification detection data every second within the standard sampling time is used as the identification detection result data and transmitted.

[0010] 3. The free acceleration method for diesel vehicles stipulates that the test is repeated three times, and the maximum value is taken as the test result data each time. The acceleration time is short, the sampling frequency is 10, and the maximum value of the identification detection data within each second is sampled as the identification detection data of that second. The maximum value of the identification detection data within the sampling time is used as the identification detection result data of each operation. The average value of the identification detection result data of the three operations is taken as the final identification detection result data and transmitted.

[0011] 4. Automobile fuel evaporation leak detection measures the fuel vapor pressure within a specified time period. The video stream of the pressure gauge recorded by the camera is collected at a sampling frequency of 5. The recognition standard stipulates that the pressure gauge values ​​at the initial and final moments of the sampling time are used as the identification test result data and transmitted.

[0012] 5. For daily calibration of exhaust gas detection equipment, use standard gas or filter with a sampling frequency of 3. Take the average value of the identification detection data within the specified sampling time as the identification calibration result data and transmit it.

[0013] The artificial intelligence supervision method for exhaust gas detection has the advantages of good accuracy, high efficiency, convenience, and automatic supervision. It can save a lot of supervision labor costs and realize uninterrupted supervision, greatly improving the authenticity, objectivity and accuracy of detection.

Claims

1. The artificial intelligence monitoring method for exhaust gas detection is a method for automatic artificial intelligence monitoring of motor vehicle exhaust emission detection, which is characterized by: The video stream of the display screen of the motor vehicle exhaust gas detection equipment is collected by a camera, and artificial intelligence image recognition technology is used to identify the detection information and data in the collected image. The identified display screen value is used as the true value and transmitted to the regulatory department; dynamic instantaneous images are obtained from the start to the end of the detection process, or dynamic instantaneous images within this period are obtained according to the two sampling start and end signals issued by the detection control system, or dynamic instantaneous images within this period are obtained from the detection process displayed by the detection equipment; the sampling frequency of the dynamic instantaneous image is in the range of 1 to 60, and the relevant detection information and data in the captured image are identified, and a single recognition detection data or the average recognition detection data within each second or the maximum recognition detection data within each second is used as the recognition detection data for that second; the recognition detection data of each dynamic instantaneous image or the recognition detection data within each second or the recognition detection result data within the data acquisition time specified by the standard are transmitted for automatic comparison with the detection process data or reported detection result data transmitted by the detection station to the regulatory department. It is suitable for artificial intelligence automatic supervision of in-use motor vehicle exhaust gas detection or daily calibration of exhaust gas detection equipment.