Signal lamp second reading method, device and equipment and readable storage medium

Abstract

A signal lamp second-reading method, apparatus and device, and a readable storage medium, the method comprising: acquiring second-reading data of each data source, including traffic management platform second-reading data, big data reckoning second-reading data, and image recognition second-reading data; carrying out confidence coefficient evaluation on the second reading data of each data source to obtain the confidence coefficient of each data source; and when the highest confidence in the confidence of each data source is greater than or equal to a confidence threshold, determining a final second-reading result according to the second-reading data of the corresponding data source, otherwise, performing weighted fusion on the second-reading data of each data source according to the confidence to obtain the final second-reading result. According to the method, the confidence degree evaluation is performed on the second-reading data of each data source, when the confidence degree meets the requirement, the data source with the highest confidence degree is used for second-reading prediction, and when the confidence degree does not meet the requirement, a plurality of data sources are synthesized according to the confidence degree for second-reading prediction, so that a relatively large prediction deviation when a single data source fails is avoided, and the accuracy of second-reading prediction is improved. And the reliability and the coverage range of the reading second prediction are effectively improved.

Description

A method, apparatus, device, and readable storage medium for counting down traffic lights. Technical Field

[0001] This application relates to the field of autonomous driving, specifically to a traffic light countdown method, apparatus, device, and readable storage medium. Background Technology

[0002] In the field of autonomous driving, traffic light countdown prediction technology is a core component for improving road traffic efficiency and driving safety. With the continuous growth of urban traffic flow and the popularization of intelligent connected vehicles, the demand for real-time and accurate traffic light countdown information is becoming increasingly urgent. There is a pressing need to achieve high-precision prediction with low error across all scenarios to support navigation systems in optimizing route planning, reducing unnecessary waiting time, and providing reliable input for autonomous driving decisions.

[0003] Currently, there are two main approaches to predicting the remaining time of traffic lights: one is to directly obtain real-time light status data transmitted by the traffic management platform through V2X communication; the other is to use machine learning algorithms to estimate the remaining time of traffic lights based on historical vehicle trajectory data (such as GPS positioning and start-stop records) passing through the intersection.

[0004] However, existing traffic light prediction methods based on a single data source all have certain shortcomings. While V2X communication methods for acquiring traffic management platform data offer high accuracy, their coverage is limited and they rely on regional infrastructure upgrades and open government policies. Methods based on historical vehicle trajectory data depend on data density and are difficult to effectively model at intersections with low traffic volume. Furthermore, these methods cannot respond promptly to temporary traffic timing adjustments in adaptive traffic light scenarios. Therefore, traffic light prediction methods based on a single data source are prone to significant deviations when the data source fails (e.g., signal interruption, model confidence fluctuations), resulting in poor prediction reliability and limited coverage, necessitating improvement. Summary of the Invention

[0005] This application provides a signal light countdown method, apparatus, device, and readable storage medium, which can solve the technical problems in the prior art that result in large deviations when the data source fails, poor prediction reliability, and limited coverage.

[0006] In a first aspect, embodiments of this application provide a traffic light countdown method, comprising: acquiring countdown data from various data sources, including countdown data from a traffic management platform, countdown data calculated from big data, and countdown data from image recognition; evaluating the confidence level of the countdown data from each data source to obtain the confidence level of each data source; when the highest confidence level among the confidence levels of each data source is greater than or equal to a confidence level threshold, determining the final countdown result based on the countdown data from the corresponding data source; otherwise, weighting and fusing the countdown data from each data source based on the confidence level to obtain the final countdown result.

[0007] In conjunction with the first aspect, in one implementation, the confidence level of the countdown data from each data source is evaluated to obtain the confidence level of each data source, including: performing a weighted evaluation on the countdown data from the traffic management platform and the countdown data from image recognition based on accuracy indicators, stability indicators, and timeliness indicators to obtain the corresponding confidence level; and obtaining the confidence level of the countdown data based on big data extrapolation through data interfaces.

[0008] In conjunction with the first aspect, in one implementation, the accuracy index of the traffic management platform's countdown data is determined based on the data error rate, the stability index is determined based on the continuous effective duration, and the timeliness index is determined based on the data refresh delay; the accuracy index of the image recognition countdown data is determined based on the target detection output confidence level, the stability index is determined based on the variance of the output result, and the timeliness index is determined based on the processing delay.

[0009] In conjunction with the first aspect, in one implementation, the weights of the weighted evaluation indicators are adjusted and optimized based on the gradient descent algorithm.

[0010] In conjunction with the first aspect, in one implementation, obtaining countdown data from each data source includes: obtaining a light color status sequence from each data source; determining countdown data from each data source based on the light color status sequence and the current timestamp, wherein the countdown data includes the current light color and the remaining seconds.

[0011] In conjunction with the first aspect, in one implementation, the traffic light countdown method further includes: determining the current actual light color based on the current light color of each data source and the data source priority; if the current light color of any data source is different from the current actual light color, setting the confidence level of that data source to zero.

[0012] In conjunction with the first aspect, in one implementation, the countdown data from each data source are weighted and fused according to the confidence level to obtain the final countdown result, including: using the confidence level as the fusion weight to weight and fuse the remaining seconds of each data source to obtain the final remaining seconds; and determining the final countdown result based on the current actual light color and the final remaining seconds.

[0013] Secondly, this application provides a traffic light countdown device, which includes: a data acquisition module for acquiring countdown data from various data sources, including countdown data from a traffic management platform, countdown data calculated from big data, and countdown data from image recognition; a confidence assessment module for assessing the confidence of the countdown data from each data source to obtain the confidence of each data source; and a countdown result determination module for determining the final countdown result based on the countdown data from the corresponding data source when the highest confidence among the confidence of each data source is greater than or equal to a confidence threshold, otherwise, the countdown data from each data source is weighted and fused based on the confidence to obtain the final countdown result.

[0014] Thirdly, this application provides a traffic light countdown device, which includes a processor, a memory, and a traffic light countdown program stored in the memory and executable by the processor. When the traffic light countdown program is executed by the processor, it implements the steps of the traffic light countdown method as described above.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing a traffic light countdown program, wherein when the traffic light countdown program is executed by a processor, it implements the steps of the traffic light countdown method as described above.

[0016] The beneficial effects of the technical solution provided in this application include: This application evaluates the confidence level of the countdown data from each data source separately, and when the confidence level meets the requirements, uses the data source with the highest confidence level to predict the countdown. When the confidence level does not meet the requirements, it combines multiple data sources based on the confidence level to predict the countdown. This avoids large prediction deviations when a single data source fails, and effectively improves the reliability and coverage of countdown prediction. Attached Figure Description

[0017] Figure 1 is a flowchart illustrating an embodiment of the traffic light countdown method of this application; Figure 2 is a flowchart illustrating the acquisition of countdown data from various data sources in an embodiment of this application; Figure 3 is a flowchart illustrating the determination of the confidence level of each data source in an embodiment of this application; Figure 4 is a flowchart illustrating the weighted fusion of countdown data in an embodiment of this application; Figure 5 is a functional module diagram illustrating an embodiment of the traffic light countdown device of this application; Figure 6 is a hardware structure diagram illustrating the traffic light countdown device involved in the embodiment of this application. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0020] Firstly, embodiments of this application provide a method for counting down the seconds of a traffic light.

[0021] In one embodiment, referring to Figure 1, which is a flowchart illustrating an embodiment of the traffic light countdown method of this application, the traffic light countdown method includes: S101, acquiring countdown data from various data sources, including countdown data from traffic management platforms, countdown data calculated from big data, and countdown data from image recognition; S102, evaluating the confidence level of the countdown data from each data source to obtain the confidence level of each data source; S103, when the highest confidence level among the confidence levels of each data source is greater than or equal to a confidence threshold, determining the final countdown result based on the countdown data from the corresponding data source; otherwise, weighting and fusing the countdown data from each data source based on the confidence level to obtain the final countdown result.

[0022] Specifically, in this embodiment, the system acquires countdown data from multiple data sources. Data acquisition includes real-time light status data obtained from a traffic management platform via V2X communication, real-time light status data derived from big data vehicle trajectory estimation via a map provider's data interface, and data obtained through image recognition of traffic light images captured by vehicle-mounted cameras or roadside units. The raw light status sequences acquired from each data source are then processed to obtain countdown data in a standard format.

[0023] To evaluate the validity of the countdown data, the implementation example assesses the confidence level of the countdown data for each data point. The countdown data from the traffic management platform and the countdown data from image recognition are obtained by weighted summation based on accuracy, stability, and timeliness indicators. The countdown data calculated from big data is obtained directly through the map provider's data interface.

[0024] Based on the determined confidence levels of each data source, if the highest confidence level is greater than or equal to the confidence threshold, the final countdown result can be directly determined based on the countdown result of the data source corresponding to the highest confidence level. If the highest confidence level is less than the confidence threshold, the embodiment determines the final countdown result by weighting and fusing the countdown results of each data source based on the obtained confidence levels.

[0025] In this embodiment, the confidence level of the countdown data from each data source is evaluated separately. When the confidence level meets the requirements, the data source with the highest confidence level is used for countdown prediction. When the confidence level does not meet the requirements, the countdown prediction is performed by combining the confidence levels of multiple data sources. This avoids large prediction deviations when a single data source fails, and effectively improves the reliability and coverage of countdown prediction.

[0026] Further, in one embodiment, Figure 2 is a flowchart illustrating the process of obtaining countdown data from each data source according to an embodiment of this application. As shown in Figure 2, obtaining countdown data from each data source includes: S201, obtaining the light color status sequence of each data source; S202, determining the countdown data of each data source based on the light color status sequence and the current timestamp. The countdown data includes the current light color and the remaining seconds.

[0027] Specifically, the implementation method obtains countdown data from three data sources: traffic management platform, big data estimation, and image recognition.

[0028] The traffic management platform data is obtained through V2X communication or direct connection to the traffic management department's signal control platform or the real-time status of traffic lights. Big data inference data is obtained by map providers offering navigation services by analyzing large amounts of anonymous vehicle trajectory data (such as start-stop patterns before stop lines) and using machine learning models to infer the signal cycle and phase change points at intersections. Image recognition data is obtained through vehicle-mounted cameras or roadside units, using computer vision algorithms to identify the color and countdown timer of traffic lights in real time.

[0029] In addition, the raw data obtained from different data sources is in the form of a light color status sequence. To facilitate subsequent weighted calculations, the embodiment calculates the result by combining the light color status sequence with the current timestamp and outputting formatted countdown data. For example, in the red light state, the output result is: {Current light color: "red", remaining seconds of red light: 15}.

[0030] In this embodiment, by unifying the data obtained from various data sources into a standard format for countdown data, subsequent processing steps can be simplified, and system errors caused by different data formats can be avoided.

[0031] Further, in one embodiment, Figure 3 is a flowchart illustrating the process of determining the confidence level of each data source according to an embodiment of this application. As shown in Figure 3, the confidence level of each data source is evaluated by assessing the countdown data of each data source, including: S301, performing a weighted evaluation on the countdown data of the traffic management platform and the countdown data of image recognition based on accuracy indicators, stability indicators and timeliness indicators to obtain the corresponding confidence level; S302, obtaining the confidence level of the countdown data based on big data estimation using the data interface.

[0032] Furthermore, in one embodiment, the accuracy index of the traffic management platform's countdown data is determined based on the data error rate, the stability index is determined based on the continuous effective duration, and the timeliness index is determined based on the data refresh delay; the accuracy index of the image recognition countdown data is determined based on the target detection output confidence level, the stability index is determined based on the variance of the output result, and the timeliness index is determined based on the processing delay.

[0033] Specifically, in the confidence assessment process, for the big data-based countdown data derived from the map provider's big data platform, this embodiment directly obtains the confidence level of the big data-based countdown data from the map provider. For the countdown data from the traffic management platform and the image recognition countdown data, this embodiment evaluates the data based on three indicators: accuracy, stability, and timeliness, and determines the confidence level through weighted merging. The weighted calculation formula is expressed as follows:

[0034] in, Indicates the first Confidence level of each data source Indicates the accuracy index. Indicates stability index, Indicators of timeliness , and The weights are the corresponding values ​​for each indicator.

[0035] For the countdown data of the traffic management platform, its data accuracy indicators This is mainly reflected in the data error rate. In this embodiment, the optional calculation method is:

[0036] in, Indicates the number of errors or exceptions. This indicates the total amount of data.

[0037] Data stability indicators of countdown data from traffic management platforms It is determined by the continuous effective duration, for example, based on the duration during which the data stream is uninterrupted (without timeouts or verification failures) within a fixed time window (e.g., 30 seconds).

[0038] Timeliness indicators of countdown data from traffic management platforms This is calculated based on the data refresh delay, and the optional calculation method is:

[0039] in, The difference (in seconds) between the current time and the data packet timestamp.

[0040] For image recognition countdown data, data accuracy indicators The confidence score output by an object detection model (such as YOLO) can be directly used as... The measurement.

[0041] The stability index of image recognition countdown data can be determined based on the variance of the recognition results, for example, based on the number of consecutive frames with consistent recognition results in a certain segment of consecutive image frames (within the same 1 second).

[0042] The calculation method for the timeliness index of image recognition countdown data is the same as that for the countdown data of the traffic management platform mentioned above. The difference is that the data refresh delay is replaced by the total delay of image acquisition, transmission and recognition.

[0043] In this embodiment, by weighting the evaluation of three indicators—accuracy, stability, and timeliness—the confidence level of the data source can be comprehensively assessed from multiple dimensions, ensuring the accuracy of the judgment.

[0044] Furthermore, in one embodiment, the traffic light countdown method further includes: determining the current actual light color based on the current light color of each data source and the data source priority; if the current light color of any data source is different from the current actual light color, setting the confidence level of that data source to zero.

[0045] Specifically, considering that a failure of a data source may result in an incorrect light color, leading to a significant deviation in the prediction results, the embodiment also adds a light color judgment mechanism, including: when the light color judgments from all data sources are consistent, that light color is used as the current actual light color; if the light color judgments from all data sources are inconsistent, the light color with the most consistent judgments is used as the current actual light color (e.g., if V2X communication and image recognition both report "red," but the map provider's big data calculation results in "green," then the result is "red"); if there is missing data (e.g., only two data sources have obtained data), the current actual light color is determined according to data priority. In the embodiment, the data priority order is: the traffic management platform's countdown data is the highest, followed by the big data calculation countdown data, and the image recognition countdown data is the lowest.

[0046] Furthermore, for data sources where the determined light color differs from the actual current light color, the implementation sets the confidence level of the data source to zero to avoid erroneous data interfering with the final result.

[0047] In this embodiment, by using the light color judgment mechanism and the confidence level zeroing mechanism, the influence of the data source on the final countdown result can be avoided when there is a serious judgment deviation in the data source (incorrect light color judgment), thereby improving accuracy.

[0048] Further, in one embodiment, Figure 4 is a flowchart of the weighted fusion of countdown data in this application embodiment. As shown in Figure 4, the countdown data from each data source is weighted and fused according to the confidence level to obtain the final countdown result, including: S401, using the confidence level as the fusion weight, the remaining seconds of each data source are weighted and fused to obtain the final remaining seconds; S402, the final countdown result is determined according to the current actual light color and the final remaining seconds.

[0049] Specifically, in the embodiment, if the highest confidence level The corresponding data source is used directly for output. If Then, a weighted fusion method is used to determine the final countdown result. The formula is expressed as:

[0050] Among them, the numerator represents the countdown results from each data source. Its confidence level The sum of the products of the data sources is used to determine the weight of the results. The higher the confidence level of a data source, the greater its weight in the final result. The denominator is the sum of the confidence levels of all data sources participating in this fusion. Its purpose is to normalize the results, ensuring that the sum of the weights of each data source is 1, so that the final result is a reasonable weighted average without being amplified or reduced.

[0051] In this embodiment, when the confidence level does not meet the requirements, the countdown prediction is performed by combining multiple data sources based on the confidence level. This avoids large prediction deviations when a single data source fails, and effectively improves the reliability and coverage of the countdown prediction.

[0052] Furthermore, in one embodiment, the weights of the weighted evaluation indicators are adjusted and optimized based on the gradient descent algorithm.

[0053] Specifically, to ensure greater accuracy in the confidence levels obtained from the confidence assessment, the embodiment also uses a gradient descent algorithm to weight each confidence level. , and Make adjustments to your learning.

[0054] In an embodiment, , and The initial values ​​can be set to 0.4, 0.3, and 0.3 respectively. The example uses real timestamps of traffic light changes over a period of time, obtained in a test environment or through independent channels, to learn and adjust the weights.

[0055] In this embodiment, the error calculation is determined by the total error between the predicted countdown value and the actual value, which can be calculated using the mean square error, expressed by the formula:

[0056] in, This represents the merged countdown prediction value. Represents the actual value. This represents the number of samples.

[0057] The gradient determines the magnitude and direction of each weight's "contribution" to the total error based on the partial derivative of the loss function with respect to each weight.

[0058] In this embodiment, the loss is reduced by adjusting the weights in the opposite direction of the gradient, and the weights are updated accordingly. The weight update formula can be expressed as:

[0059] in, This indicates the updated weights. This indicates the weights before the update. This is the learning rate.

[0060] The weight optimization is completed by iterating through the entire update process until the loss function converges or the predetermined number of iterations is reached.

[0061] In this embodiment, by iteratively learning and optimizing the weights of each confidence level during the confidence level calculation process, the accuracy of the confidence level calculation can be further improved, thereby ensuring the reliability of the traffic light countdown results.

[0062] Secondly, embodiments of this application also provide a signal light countdown device.

[0063] In one embodiment, referring to Figure 5, which is a functional block diagram of a traffic light countdown device according to an embodiment of the present application, the traffic light countdown device includes: a data acquisition module 501, used to acquire countdown data from various data sources, including countdown data from traffic management platforms, countdown data calculated from big data, and countdown data from image recognition; a confidence assessment module 502, used to assess the confidence of the countdown data from each data source to obtain the confidence of each data source; and a countdown result determination module 503, used to determine the final countdown result based on the countdown data from the corresponding data source when the highest confidence among the confidence of each data source is greater than or equal to the confidence threshold; otherwise, the countdown data from each data source is weighted and fused based on the confidence to obtain the final countdown result.

[0064] Furthermore, in one embodiment, the confidence assessment module is used to: perform a weighted assessment of the countdown data from the traffic management platform and the countdown data from image recognition based on accuracy indicators, stability indicators, and timeliness indicators to obtain the corresponding confidence level; and obtain the confidence level of the countdown data based on big data estimation through the data interface.

[0065] Furthermore, in one embodiment, in the confidence assessment module, the accuracy index of the traffic management platform's countdown data is determined based on the data error rate, the stability index is determined based on the continuous effective duration, and the timeliness index is determined based on the data refresh delay; the accuracy index of the image recognition countdown data is determined based on the target detection output confidence, the stability index is determined based on the variance of the output result, and the timeliness index is determined based on the processing delay.

[0066] Furthermore, in one embodiment, in the confidence assessment module, the weights of the weighted assessment indicators are adjusted and optimized based on the gradient descent algorithm.

[0067] Furthermore, in one embodiment, the data acquisition module is used to: acquire the light color status sequence of each data source; and determine the countdown data of each data source based on the light color status sequence and the current timestamp, wherein the countdown data includes the current light color and the remaining seconds.

[0068] Furthermore, in one embodiment, the traffic light countdown device further includes a new module for: determining the current actual light color based on the current light color of each data source and the data source priority; and setting the confidence level of any data source to zero if the current light color of any data source is different from the current actual light color.

[0069] Furthermore, in one embodiment, the countdown result determination module is used to: use the confidence level as the fusion weight to perform weighted fusion of the remaining seconds of each data source to obtain the final remaining seconds; and determine the final countdown result based on the current actual light color and the final remaining seconds.

[0070] The functions of each module in the above-mentioned traffic light countdown device correspond to the steps in the above-mentioned traffic light countdown method embodiment, and their functions and implementation processes will not be described in detail here.

[0071] Thirdly, this application provides a traffic light countdown device, which can be a device with data processing capabilities, such as an on-board ECU (Electronic Control Unit).

[0072] Referring to Figure 6, which is a schematic diagram of the hardware structure of the traffic light countdown device involved in the embodiment of this application, the traffic light countdown device may include a processor, a memory, a communication interface, and a communication bus in this embodiment.

[0073] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0074] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting internal components of the traffic light countdown device, as well as interfaces used for interconnecting the traffic light countdown device with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0075] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0076] The processor can be a general-purpose processor, which can call the traffic light countdown program stored in the memory and execute the traffic light countdown method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the traffic light countdown program is called can be referred to in the various embodiments of the traffic light countdown method of this application, and will not be repeated here.

[0077] Those skilled in the art will understand that the hardware structure shown in FIG6 does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0078] Fourthly, embodiments of this application also provide a computer-readable storage medium.

[0079] The present application provides a computer-readable storage medium storing a traffic light countdown program, wherein when the traffic light countdown program is executed by a processor, it implements the steps of the traffic light countdown method described above.

[0080] The method implemented when the traffic light countdown program is executed can be referred to in various embodiments of the traffic light countdown method of this application, and will not be repeated here.

[0081] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0082] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0083] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0084] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0085] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0086] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods of the various embodiments of this application.

[0087] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for counting down seconds on a traffic light, characterized in that, include: The system acquires countdown data from various data sources, including countdown data from traffic management platforms, countdown data calculated from big data, and countdown data from image recognition. It then evaluates the confidence level of each data source's countdown data to obtain the confidence level of each data source. If the highest confidence level among all data sources is greater than or equal to a confidence threshold, the final countdown result is determined based on the countdown data from the corresponding data source. Otherwise, the countdown data from each data source is weighted and fused based on the confidence level to obtain the final countdown result.

2. The traffic light countdown method according to claim 1, characterized in that, The confidence assessment of the countdown data from each data source is performed to obtain the confidence level of each data source, including: performing a weighted assessment of the countdown data from the traffic management platform and the countdown data from the image recognition based on accuracy indicators, stability indicators, and timeliness indicators to obtain the corresponding confidence level; and obtaining the confidence level of the countdown data calculated from the big data based on the data interface.

3. The traffic light countdown method according to claim 2, characterized in that, The accuracy index of the traffic management platform's countdown data is determined based on the data error rate, the stability index is determined based on the continuous effective duration, and the timeliness index is determined based on the data refresh delay; the accuracy index of the image recognition countdown data is determined based on the target detection output confidence level, the stability index is determined based on the variance of the output result, and the timeliness index is determined based on the processing delay.

4. The traffic light countdown method according to claim 2, characterized in that, The weights of the indicators in the weighted evaluation are adjusted and optimized based on the gradient descent algorithm.

5. The traffic light countdown method according to claim 1, characterized in that, The step of obtaining countdown data from each data source includes: obtaining the light color status sequence of each data source; determining the countdown data of each data source based on the light color status sequence and the current timestamp, wherein the countdown data includes the current light color and the remaining seconds.

6. The traffic light countdown method according to claim 5, characterized in that, The method further includes: determining the current actual light color based on the current light color of each data source and the data source priority; if the current light color of any data source is different from the current actual light color, setting the confidence level of that data source to zero.

7. The traffic light countdown method according to claim 6, characterized in that, The step of weighting and fusing the countdown data from each data source based on the confidence level to obtain the final countdown result includes: using the confidence level as the fusion weight to weight and fuse the remaining seconds from each data source to obtain the final remaining seconds; and determining the final countdown result based on the current actual light color and the final remaining seconds.

8. A signal light countdown device, characterized in that, The traffic light countdown device includes: a data acquisition module for acquiring countdown data from various data sources, including countdown data from traffic management platforms, countdown data calculated from big data, and countdown data from image recognition; a confidence assessment module for assessing the confidence of the countdown data from each data source to obtain the confidence of each data source; and a countdown result determination module for determining the final countdown result based on the countdown data from the corresponding data source when the highest confidence among the confidence values ​​from each data source is greater than or equal to a confidence threshold; otherwise, the countdown data from each data source is weighted and fused based on the confidence values ​​to obtain the final countdown result.

9. A signal light countdown device, characterized in that, The traffic light countdown device includes a processor, a memory, and a traffic light countdown program stored in the memory and executable by the processor, wherein when the traffic light countdown program is executed by the processor, it implements the steps of the traffic light countdown method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a traffic light countdown program, wherein when the traffic light countdown program is executed by a processor, it implements the steps of the traffic light countdown method as described in any one of claims 1 to 7.

Images