Laser radar ranging method, laser radar, and ranging correction device

Abstract

The application discloses a laser radar ranging method, a laser radar and a ranging correction device. The method comprises the following steps: acquiring ranging data when the laser radar is working; determining that the original ranging result needs to be corrected according to a noise value; determining an error correction section to which the ranging data belongs according to echo characteristics, and selecting an error correction parameter corresponding to the error correction section; substituting the error correction parameter into a preset laser radar ranging correction model to determine a corresponding ranging error correction function; inputting the noise value into the ranging error correction function to obtain a distance error value; and correcting the original ranging result according to the distance error value to obtain a target ranging result of the laser radar. Without pre-identifying the target material or the environmental type, the application effectively reduces the influence of noise and echo characteristic differences on the ranging result, makes the corrected ranging result closer to the real distance, and improves the ranging precision of the laser radar.

Description

Technical Field

[0001] This application relates to the field of lidar technology, and in particular to a lidar ranging method, lidar, and ranging correction device. Background Technology

[0002] The spectral characteristics of sunlight are quite complex, and its spectral range covers the operating band of lidar. When lidar is in an outdoor environment or is exposed to sunlight indoors, it is easy to introduce external noise in the same frequency band as the effective echo signal, which will cause the lidar to deviate in ranging and affect the normal operation of lidar. Summary of the Invention

[0003] One objective of this application is to provide a lidar ranging method, lidar, and ranging correction device to solve the technical problem of how to improve the ranging accuracy of lidar.

[0004] To address the aforementioned technical problems, one technical solution adopted in this application is as follows: a lidar ranging method is provided, comprising: acquiring ranging data when the lidar is operating, wherein the ranging data includes noise values, echo characteristics, and original ranging results; when it is determined based on the noise values ​​that the original ranging results need to be corrected, determining the error correction segment to which the ranging data belongs based on the echo characteristics, and selecting the error correction parameter corresponding to the error correction segment; substituting the error correction parameter into a preset lidar ranging correction model to determine the corresponding ranging error correction function; inputting the noise values ​​into the ranging error correction function to calculate the distance error value; and correcting the original ranging results based on the distance error value to obtain the target ranging result of the lidar.

[0005] In some embodiments, when it is determined that the original ranging result needs to be corrected based on the noise value, the error correction segment to which the ranging data belongs is determined based on the echo characteristics, and the error correction parameter corresponding to the error correction segment is selected, including: comparing the noise value with a preset noise threshold; if the noise value is greater than the preset noise threshold, it is determined that the original ranging result needs to be corrected; when it is determined that the original ranging result needs to be corrected, comparing the echo characteristics with at least one preset echo characteristic threshold, and dividing the ranging data into the corresponding error correction segment based on the comparison result; and selecting the error correction parameter corresponding to the error correction segment from a pre-calibrated set of error correction parameters according to the error correction segment to which the ranging data belongs.

[0006] In some embodiments, the lidar includes a photosensitive chip, which includes a first region and a second region. The first region is a normal receiving region for receiving echo signals and outputting noise values, and the second region is a weakened receiving region for performing energy attenuation processing on the received echo signals and outputting echo peak values. When the lidar is working, ranging data is acquired, including: acquiring the noise value output by the first region; acquiring the echo peak value output by the second region; and acquiring the original ranging result corresponding to the noise value and the echo peak value.

[0007] In some embodiments, correcting the original ranging result based on the distance error value to obtain the target ranging result of the lidar includes: performing a subtraction operation between the original ranging result and the distance error value, and using the difference as the target ranging result of the lidar.

[0008] In some embodiments, the method further includes: obtaining an error correction parameter set; obtaining the error correction parameter set includes: obtaining multiple sets of calibration ranging data, the calibration ranging data including noise values, echo characteristics, original ranging results, and true distance; calculating the ranging error based on the original ranging results and true distance; segmenting the multiple sets of calibration ranging data based on noise values ​​and echo characteristics to obtain multiple error correction segments; extracting the ranging error of each error correction segment; fitting the ranging error to each error correction segment using a preset lidar ranging correction model to obtain the error correction parameters corresponding to each error correction segment; wherein the error correction parameters corresponding to each error correction segment together constitute the error correction parameter set.

[0009] In some embodiments, the lidar is mounted on a calibration fixture, and a calibration target is positioned at a preset distance in front of the lidar. The calibration target includes objects with different reflectivity, including at least low-reflectivity, medium-reflectivity, and high-reflectivity targets. A light source is positioned within a preset range of the calibration target, and the light source is used to illuminate the calibration target with different light intensities. Multiple sets of calibration ranging data are acquired, including: controlling the actual distance between the lidar and the calibration target, and collecting noise values, echo characteristics, and original ranging results at multiple different actual distances, under different target objects, and under different light intensities. The noise values, echo characteristics, and original ranging results acquired in the same ranging event collectively correspond to the actual distance, which is used as the true distance.

[0010] In some embodiments, multiple sets of calibration ranging data are segmented based on noise values ​​and echo characteristics to obtain multiple error correction segments. This includes: statistically analyzing the correspondence between noise values ​​and ranging errors in multiple sets of calibration ranging data to obtain the distribution of noise values ​​and ranging errors; determining at least one noise threshold based on the distribution of noise values ​​and ranging errors, and performing preliminary segmentation of multiple sets of calibration ranging data based on the noise threshold; after completing the preliminary segmentation, analyzing the correspondence between echo characteristics and ranging errors in each preliminary segment to obtain the distribution of echo characteristics and ranging errors; determining at least one echo characteristic threshold based on the distribution of echo characteristics and ranging errors, and further subdividing the preliminary segmentation based on the echo characteristic threshold to obtain multiple error correction segments.

[0011] To solve the above-mentioned technical problems, one technical solution adopted in this application is to provide a laser radar, including a memory and a processor. The memory is connected to the processor, and the processor is used to execute one or more computer programs stored in the memory. When the processor executes one or more computer programs, it enables the laser radar to implement the laser radar ranging method described above.

[0012] To address the aforementioned technical problems, one technical solution adopted in this application is to provide a ranging correction device, comprising: a memory and a processor. The memory is connected to the processor, which receives ranging data output by a lidar and executes one or more computer programs stored in the memory based on the ranging data. When the processor executes the one or more computer programs, it enables the ranging correction device to implement the lidar ranging method described above. The ranging correction device includes, but is not limited to, a signal processing module integrated within the lidar, a lidar main control chip, a microcontroller unit, a digital signal processor, a system-on-a-chip, a lidar control board, a lidar sensor module, or an external processing unit, edge computing device, vehicle controller, robot controller, industrial controller, or server that is communicatively connected to the lidar.

[0013] To solve the above-mentioned technical problems, one technical solution adopted in the embodiments of this application is: providing a non-volatile computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by the lidar, the lidar performs the lidar ranging method as described above; when the computer-executable instructions are executed by the ranging correction device, the ranging correction device performs the lidar ranging method as described above.

[0014] The lidar ranging method, lidar, and ranging correction device provided in this application, during the actual operation of the lidar, acquire ranging data in real time, including noise values, echo characteristics, and the original ranging result. First, the reliability of the ranging data is judged using the noise value. Error correction is triggered only when correction of the original ranging result is needed, thus avoiding over-processing of low-noise, stable ranging data and improving the overall stability of the system. When correction is required, the ranging data is further segmented based on echo characteristics, and corresponding error correction parameters are matched for different segments, enabling the ranging error correction process to adapt to target scenes with different echo intensities and reflection characteristics. By substituting the selected error correction parameters into a preset ranging error correction model and calculating the distance error value based on the noise value, correction of the current ranging result is achieved. Therefore, this solution effectively reduces the impact of noise and echo characteristic differences on the ranging result without prior identification of the target material or environment type, making the corrected ranging result closer to the true distance and improving the ranging accuracy of the lidar. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of this application;

[0017] Figure 2 This is a flowchart illustrating a method for obtaining a set of error correction parameters provided in an embodiment of this application;

[0018] Figure 3 This is a schematic diagram illustrating the relationship between noise value and error value provided in the embodiments of this application;

[0019] Figure 4a , Figure 4b , Figure 4c and Figure 4d This is a schematic diagram illustrating the relationship between the actual distance and the noise value provided in the embodiments of this application;

[0020] Figure 5 This is a schematic diagram illustrating the relationship between real distance and echo characteristics provided in an embodiment of this application;

[0021] Figure 6 This is a schematic diagram showing the relationship between noise value and ranging error provided in the embodiments of this application;

[0022] Figure 7This is a flowchart of a lidar ranging method provided in an embodiment of this application;

[0023] Figure 8 This is a schematic diagram of the structure of a lidar provided in an embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0025] It should be noted that, unless there is a conflict, the various features in the embodiments of this application can be combined with each other, all of which are within the protection scope of this application. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. Moreover, the terms "first" and "second" used in this application do not limit the data or execution order, but only distinguish between identical or similar items with essentially the same function and effect.

[0026] Sunlight has a continuous and wide-spectrum frequency characteristic, containing spectral components that overlap with the emission band of lidar. When lidar operates outdoors, or is exposed to direct, reflected, or diffuse sunlight indoors, the strong light components in the ambient light enter the lidar's receiving channel. These components, along with the effective echo signal emitted and returned by the lidar, are simultaneously received by the lidar's photosensitive chip, thus creating external noise interference at the receiving end that operates in the same frequency band as the effective signal.

[0027] In the aforementioned scenarios, strong external light not only significantly raises the noise baseline of the received signal and reduces the signal-to-noise ratio, but may also cause energy saturation in some pixels or receiving areas of the photosensitive chip, resulting in distortions in characteristics such as the amplitude and time of arrival of the echo signal. Especially under high illumination conditions or in the presence of highly reflective targets, this type of noise interference is further amplified, easily causing time-of-flight (TOF) detection errors, leading to deviations or even failures in the ranging results output by the lidar. This affects the reliability and robustness of lidar in applications such as outdoor ranging, autonomous driving, robotic perception, and industrial inspection. Therefore, a solution is urgently needed to improve the ranging accuracy and stability of lidar in complex lighting environments under conditions of ambient light interference.

[0028] Therefore, this application provides a lidar ranging method. During actual lidar operation, this method acquires ranging data in real-time, including noise levels, echo characteristics, and the original ranging result. First, it uses the noise level to assess the reliability of the ranging data. Error correction is triggered only when the original ranging result needs correction, thus avoiding over-processing of low-noise, stable ranging data and improving the overall system stability. When correction is needed, the ranging data is further segmented based on echo characteristics, and corresponding error correction parameters are matched to different segments. This allows the ranging error correction process to adapt to target scenes with different echo intensities and reflection characteristics. By substituting the selected error correction parameters into a preset ranging error correction model and calculating the distance error based on the noise level, correction of the current ranging result is achieved. Thus, this lidar ranging method effectively reduces the impact of noise and echo characteristic differences on the ranging result without requiring prior identification of the target material or environment type, making the corrected ranging result closer to the true distance and improving the ranging accuracy of the lidar.

[0029] The lidar ranging method provided in this application mainly includes two stages: a data calibration stage and a ranging method application stage.

[0030] During the data calibration phase, ranging data is acquired and analyzed under controlled illumination conditions using lidar to obtain ranging data under different distances, illuminance levels, and target reflection characteristics. Specifically, multiple lidars are sequentially arranged on a calibration fixture, with a target reflector of a predetermined material placed in front of them in the ranging direction, and ambient illumination conditions of different intensities are simulated using an external light source. During this process, the lidar outputs ranging-related data, including time-of-flight, echo energy, and noise characteristics. Based on the acquired data samples and the correlation between noise characteristics and ranging errors, the ranging error is modeled and analyzed. The model type and corresponding parameters of the lidar ranging correction model are obtained through fitting, thus completing the calibration of the ranging correction model.

[0031] In the application phase of the ranging method, the lidar ranging correction model and its related parameters obtained during the data calibration phase are stored in the lidar or its communication-connected processing unit. When the lidar performs ranging in the actual working environment, the ranging data output by the lidar is acquired in real time. The ranging data includes at least the original ranging result, noise value, and echo characteristics. The noise value is used to determine whether the current ranging data needs correction. If the noise value indicates that correction is needed, the error correction segment to which the ranging data belongs is determined based on the echo characteristics in the ranging data, and the error correction parameter corresponding to the error correction segment is selected from a pre-stored set of error correction parameters. Subsequently, the error correction parameter is substituted into the preset lidar ranging correction model to determine the ranging error correction function corresponding to the ranging data. The real-time noise value is used as input to this ranging error correction function to calculate the corresponding distance error value. Finally, the original ranging result is corrected based on the distance error value to obtain the corrected target ranging result, thereby improving the ranging accuracy and stability of the lidar in complex lighting environments.

[0032] The two stages described above are explained in detail below through specific embodiments.

[0033] Please see Figure 1 , Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of this application. This application scenario can be used for data calibration. The application scenario includes a calibration fixture 11, a lidar 12, a calibration target 13, and a light source 14. The lidar 12 is mounted on the calibration fixture 11, and the calibration target 13 is positioned at a predetermined distance in front of the lidar 12. The calibration target 13 includes targets with different reflectivity characteristics, including at least low-reflectivity targets, medium-reflectivity targets, and high-reflectivity targets. The light source 14 is positioned within a predetermined range of the calibration target 13 and is used to illuminate the calibration target 13 with different light intensities.

[0034] Among them, low-reflectivity targets are targets with low reflectivity to the light emitted by lidar 12, used to simulate ranging scenarios in low-reflectivity environments. Medium-reflectivity targets are targets with moderate reflectivity to the light emitted by lidar 12, used to simulate ranging scenarios for conventional targets. High-reflectivity targets are targets with high reflectivity to the light emitted by lidar 12, used to simulate ranging scenarios with high reflectivity or prone to echo energy saturation.

[0035] In one specific embodiment, a low-reflectivity target may include black cardboard or a target with a low-reflectivity coating, a medium-reflectivity target may include white cardboard or a target with a medium-reflectivity surface, and a high-reflectivity target may include a target made of a high-reflectivity material, such as a reflective film, a lattice structure target, or other high-reflectivity medium, but the embodiments of this application are not limited thereto.

[0036] During the calibration process, multiple different ranging conditions are constructed by adjusting the relative distance between the lidar 12 and the calibration target 13 to collect ranging data under different distance conditions. In one embodiment, the calibration distance includes a near-range segment and a far-range segment, wherein a smaller distance step size is used for data acquisition in the near-range segment and a larger distance step size is used for data acquisition in the far-range segment. For example, in a near-range range of approximately 1 m, the distance step size can be set to approximately 100 mm; in a distance range of approximately 1 m to 8 m, the distance step size can be set to approximately 1 m, so as to improve calibration efficiency while ensuring calibration accuracy.

[0037] Simultaneously, by adjusting the light intensity of the light source 14, multiple different illuminance levels are set at each calibrated distance to simulate ranging scenarios under different ambient lighting conditions. In one embodiment, the illuminance levels cover an illuminance range of approximately 0 kLux to approximately 20 kLux, and can be set in illuminance steps of approximately 1 kLux to approximately 3 kLux, thereby collecting ranging-related data output by the lidar 12 under different distances, targets with different reflectivity characteristics, and different illuminance conditions.

[0038] After the calibration environment is set up as described above, the calibration process for LiDAR 12 begins. During this calibration process, LiDAR 12 is controlled to perform ranging operations sequentially under different calibration distances, different target reflection characteristics, and different light intensities, while real-time acquisition of ranging-related data output by LiDAR 12. The acquired data serves as the basis for subsequent ranging error analysis and the establishment of a ranging correction model, characterizing the impact of changes in ambient light and target reflection characteristics on the ranging results of LiDAR 12.

[0039] In the calibration process of lidar 12, the main purpose is to obtain a set of error correction parameters for subsequent ranging correction. This set of error correction parameters includes error correction parameters corresponding to different error correction segments. Each error correction parameter is used to characterize the error relationship between the ranging data output by lidar 12 and the actual ranging result under specific ranging conditions.

[0040] Specifically, please refer to Figure 2 , Figure 2 This is a flowchart illustrating a method for obtaining a set of error correction parameters according to an embodiment of this application. The method may include the following steps:

[0041] S21. Acquire multiple sets of calibration ranging data, including noise values, echo characteristics, original ranging results, and actual distances.

[0042] The calibration ranging data is a set of data obtained by the lidar in a preset calibration environment when ranging targets. This data set is used to characterize the ranging performance of the lidar under different distances, target reflection characteristics, and lighting conditions. For example, based on... Figure 1 The application scenario shown is used to obtain the calibration ranging data.

[0043] The noise value represents the intensity of environmental interference or system noise in the signal received by the lidar during ranging, excluding the effective echo. In this embodiment, the noise value is mainly affected by sunlight. In the calibration environment, different ambient lighting conditions, including different light intensities and illuminance levels, are simulated using light source 14 to obtain the noise value distribution of the lidar under low-light, medium-light, and strong-light environments. By adjusting the illuminance of light source 14, noise values ​​corresponding to different lighting conditions can be systematically collected. This noise value is used to determine the reliability of the ranging data and to assist in determining whether the original ranging results need to be corrected.

[0044] Echo characteristics are used to describe the properties of target echo signals received by lidar, and may include information such as echo intensity and echo peak. Echo characteristics can reflect the reflectivity of a target; for example, low-reflectivity targets have weak echoes, while high-reflectivity targets have strong echoes.

[0045] The raw ranging result refers to the uncorrected direct ranging output of the lidar, typically calculated by the lidar based on time-of-flight (TOF). The raw ranging result may be affected by factors such as noise, target reflection characteristics, and ambient lighting, resulting in errors. The true distance refers to the actual physical distance between the calibration target and the lidar during lidar calibration, which can be predetermined using precision measuring tools or calibration fixtures.

[0046] In this embodiment, the photosensitive chip of the lidar is divided into two regions, both of which can detect return signals. Due to the reflective characteristics of certain target materials, the echo energy may be high, easily leading to energy saturation in some areas of the photosensitive chip. To avoid energy saturation, one of the two regions is specially processed to weaken the received echo energy, thereby ensuring the reliability of the ranging data. For example, the second region is processed to be a weakened receiving region, where the received echo signal undergoes energy attenuation processing to output the echo peak value. The first region, on the other hand, is a normal receiving region, used to receive the echo signal and output the noise value.

[0047] After processing, the normal receiving area can output complete ranging data, including time of flight (TOF), peak echo, and noise. The weakened receiving area can also output corresponding data, but for centralized processing and simplified calculation, only the peak echo and noise values ​​can be retained. For ease of differentiation and analysis, the data output from the normal receiving area are denoted as TOF, Peak1, and Noise1, respectively, and the data output from the weakened receiving area are denoted as Peak2 and Noise2, respectively.

[0048] In the calibration process of this embodiment, for the analysis of ranging error and the establishment of the model, the noise value Noise1 and the echo peak value Peak2 are selected as key features to divide the error correction segments and fit the ranging correction model, so as to provide data support for subsequent ranging correction.

[0049] The selection of noise value (Noise1) and echo peak value (Peak2) as key features is based on their ability to effectively reflect the relationship between lidar ranging error and external lighting and target reflection characteristics. Noise1 reflects the interference intensity of ambient light (such as sunlight) on the ranging signal and can be used to determine whether the original ranging result needs correction. Echo peak value (Peak2) characterizes signal saturation that may be caused by highly reflective targets and is used to divide error correction segments and select corresponding error correction parameters. In contrast, echo peak value (Peak1) has poor numerical stability due to potential signal saturation or strong reflection, making it unsuitable as a basis for segmentation or input into the model. Time-of-flight (TOF) is only used for the original ranging calculation, and its numerical value's relationship with the error trend may not be as intuitive as that of Noise1 and Peak2. Therefore, combining noise value (Noise1) and echo peak value (Peak2) in weakened regions can more reliably characterize the ranging error characteristics, thereby improving the accuracy and stability of the ranging correction model.

[0050] Once the calibration environment is set up, lidar calibration data can be acquired. This involves mounting and securing the lidar on a calibration fixture, placing calibration targets with different reflection characteristics (low, medium, and high reflectivity) at preset distances, setting the light source position and initial illumination intensity, selecting an appropriate illumination range and step size, confirming all calibration system parameters and safety status, and initiating the automatic data acquisition program. These processes can be manually assisted. After the automatic acquisition program starts, the lidar can continuously acquire calibration ranging data under different distances, illumination conditions, and target reflections. The calibration process can be performed sequentially or simultaneously on multiple lidars of different types, including devices with different models, transmit powers, or receive sensitivities. By acquiring and analyzing calibration ranging data from different types of lidars, the ranging error characteristics of various lidars in complex environments can be obtained, thereby improving the reliability of the calibration results.

[0051] Through the above process, the lidar continuously outputs calibration ranging data under different distances, light intensities, and target reflection characteristics. Each set of calibration ranging data includes at least the noise value (e.g., Noise1), echo characteristics (e.g., Peak2), the original ranging result, and the actual distance, thus obtaining multiple sets of calibration ranging data. In this embodiment, the peak echo 2 output from the weakened receiving area is used as an echo characteristic.

[0052] S22. Calculate the distance measurement error based on the original distance measurement results and the actual distance.

[0053] The ranging error is obtained by subtracting the original ranging result from the true distance, thus creating a ranging error dataset. The ranging error can be analyzed in conjunction with noise values; for example, by selecting the noise value dataset corresponding to Noise1 and the ranging error dataset, a dataset can be created... Figure 3 The diagram showing the relationship between noise value and error value is shown below. Figure 3 In the figure, the horizontal axis represents the noise value, specifically the noise value Noise1 output from the first detection area, and the vertical axis represents the corresponding ranging error (DisErr), i.e., the error value. The red dots in the figure represent data collected when the calibrated target 13 is a high-reflectivity target (such as a 3M card or a lattice card), the blue dots represent data collected when the calibrated target 13 is a medium-reflectivity target (white card), and the black dots represent data collected when the calibrated target 13 is a low-reflectivity target (black card).

[0054] Depend on Figure 3It can be seen that Noise1 and DisErr exhibit significantly different distribution characteristics under different target reflectivity and illumination conditions. If all the above data are used indiscriminately to build a Noise1-DisErr fitting model, the fitting effect is poor. The main reason is that the error variation patterns corresponding to targets with different reflectivity characteristics are significantly different, resulting in a relatively scattered distribution of data points, making it difficult to accurately describe the relationship between various types of data using a single model.

[0055] Therefore, it is necessary to segment or classify the calibration data based on echo characteristics, and establish corresponding ranging error correction models in different data segments to improve the accuracy of error fitting and the effectiveness of subsequent ranging correction.

[0056] S23. Based on noise values ​​and echo characteristics, multiple sets of calibration ranging data are segmented to obtain multiple error correction segments.

[0057] The error correction segmentation refers to the intervals in which the ranging error has a relatively consistent variation pattern. These intervals are obtained by dividing multiple sets of calibration ranging data based on noise values ​​and echo characteristics.

[0058] Specifically, the process of obtaining the above error correction segments may include:

[0059] S231. Statistically analyze the correspondence between noise values ​​and ranging errors in multiple sets of calibration ranging data to obtain the distribution of noise values ​​and ranging errors.

[0060] For details on the distribution of noise values ​​and ranging errors, please refer to [link / reference]. Figure 3 .

[0061] S232. Based on the distribution of noise values ​​and ranging errors, determine at least one noise threshold, and perform preliminary segmentation of multiple sets of calibration ranging data according to the noise threshold.

[0062] Based on the distribution of noise values ​​and ranging errors, the changing trends and dispersion of ranging errors within different noise intervals can be analyzed to determine at least one noise threshold for characterizing significant changes in ranging errors. When the noise value is below this threshold, the overall ranging error is small and concentrated, and the ranging data within this noise interval can be considered to have little impact on the ranging result and does not require correction. When the noise value is greater than or equal to the threshold, the ranging error increases significantly and exhibits different changing characteristics. In this case, the ranging data within this noise interval has a greater impact on the ranging result and requires correction. Therefore, based on the determined noise threshold, multiple sets of calibration ranging data can be initially segmented according to the intervals where the noise values ​​are located to distinguish between calibration ranging data that does not require correction and calibration ranging data that requires further error correction processing. Thus, the result of this initial segmentation includes both calibration ranging data that does not require correction and calibration ranging data that requires further error correction processing.

[0063] Among them, it is possible to establish such as Figures 4a to 4d The distribution of the relationship between the true distance and the noise value Noise1 is shown below. Figure 4a This represents the data distribution when the black scatter dots correspond to a low-reflectivity target (black card). Figure 4b The blue scatter dots represent the data distribution when the calibrated target is a medium-reflectivity target (white card). Figure 4c The red scatter dots represent the data distribution when the calibrated target is a highly reflective target (such as a 3M card or a lattice card). Figure 4d It is a data distribution formed by superimposing mixed-color scatter points corresponding to targets with different reflectivity under specific distance and lighting conditions. Among them, Figures 4a to 4d The horizontal axis represents the actual distance, and the vertical axis represents the noise value (Noise1).

[0064] based on Figures 4a to 4d The noise distribution shown can determine the noise thresholds for various calibration targets, which are used for preliminary segmentation and error correction. For example, for a medium-reflectivity target (white card), a noise value Noise1 greater than the threshold wNoise1Thr (e.g., wNoise1Thr=1300) can be set to enter the error correction segment; for a high-reflectivity target (3M card), a noise value greater than the threshold cNoise1Thr (e.g., cNoise1Thr=600) can be set to enter the error correction segment; for a black card and the area where white and black cards merge, a noise value between bNoise1Thr and wNoise1Thr (e.g., bNoise1Thr=500, wNoise1Thr=1300) can be set to enter the corresponding correction segment.

[0065] By using the threshold division described above, preliminary data segmentation based on noise characteristics can be achieved, providing a foundation for subsequent echo characteristic analysis and error correction model establishment.

[0066] S233. After completing the initial segmentation, analyze the correspondence between echo characteristics and ranging errors in each initial segment to obtain the distribution of echo characteristics and ranging errors.

[0067] After completing the initial segmentation based on the noise value Noise1, the calibration ranging data within each initial segment are further analyzed to determine the correspondence between the echo characteristics constructed based on the weakened receiving region echo peak Peak2 and the ranging error DisErr.

[0068] First, establish such as Figure 5 The diagram shows the relationship between the true distance (distance) and the echo peak value (Peak2), where the horizontal axis represents the true distance (distance) and the vertical axis represents the echo peak value (Peak2). The distribution includes red, blue, and black scatter points, which correspond to the calibration ranging data collected when the calibrated target is a high-reflectivity target (such as a 3M card or lattice card), a medium-reflectivity target (white card), and a low-reflectivity target (black card), respectively.

[0069] Figure 5 The relationship between the true distance and the peak echo value Peak2 shown is used to observe the differences in Peak2 for different reflecting targets at various true distances and to determine the basis for subdivision. Subsequently, within each preliminary segment, the data is grouped according to Peak2, and the ranging error DisErr corresponding to each group is statistically analyzed to obtain the distribution of echo characteristics and ranging error, providing a basis for determining the echo characteristic threshold and establishing error correction segments in the following steps.

[0070] S234. Based on the distribution of echo characteristics and ranging errors, determine at least one echo characteristic threshold, and further subdivide the preliminary segmentation based on the echo characteristic threshold to obtain multiple error correction segments.

[0071] Through observation Figure 5It can be observed that even within the same initial segment, there are significant differences in Peak2 values ​​for different reflective targets, providing a reference for determining the echo characteristic threshold. Based on this analysis, an echo characteristic threshold can be set to further subdivide the data within the initial segment into different error correction segments. For example, for medium reflective targets (white cards), Peak2 can be set to be less than the threshold cPeak2Thr (e.g., cPeak2Thr=1200); for high reflective targets (3M cards), Peak2 can be set to be greater than the threshold cPeak2Thr; for low reflective targets and the area where black and white cards merge, Peak2 can be set to be less than the threshold cPeak2Thr. Through this echo characteristic threshold division, fine-grained grouping of the initial segmented data can be achieved, providing a basis for subsequently establishing a segmentation error correction model.

[0072] For example, such as Figure 6 As shown, Figure 6 This is a schematic diagram showing the relationship between the noise value Noise1 and the ranging error DisErr, where the horizontal axis represents the noise value Noise1 and the vertical axis represents the ranging error DisErr. Figure 6 The data points in the data are derived from data that has already undergone preliminary segmentation (based on Noise1, see [link]). Figures 4a-4d ) and echo feature subdivision (based on Peak2, see Figure 5 The calibrated ranging data were filtered. Each data point belongs to a specific error correction segment, and the correspondence between Noise1 and DisErr reflects the impact of noise level on ranging error within that error correction segment. Through analysis... Figure 6 The data distribution of each error correction segment allows for the establishment of primary or secondary ranging error correction models within each segment. This provides a basis for subsequent application stages to select segments and calculate ranging corrections based on noise values ​​and echo characteristics, thereby improving the ranging accuracy of lidar under complex lighting conditions. For example, Figure 6 As shown, the low-noise segments without straight lines are segments that do not require processing. The data points corresponding to the three straight lines are error correction segments that need correction. The black scattered points near the black lines correspond to the calibration ranging data of low-reflectivity targets (such as black cards), the blue scattered points near the blue lines correspond to the calibration ranging data of medium-reflectivity targets (such as white cards), and the red scattered points near the red lines correspond to the calibration ranging data of high-reflectivity targets (such as 3M cards or lattice cards). Each straight line reflects the fitting relationship between the noise value and the ranging error within that segment, and is used to establish the corresponding ranging error correction model, thereby achieving segmented correction for targets with different reflection characteristics and ensuring that the corrected ranging results are more accurate and stable.

[0073] S24. Extract the ranging error of each error correction segment.

[0074] Based on Figure 6Extraction. For example, for a white card, select data points where Noise1 is greater than the threshold wNoise1Thr (e.g., wNoise1Thr=1300) and Peak2 is less than the threshold cPeak2Thr (e.g., cPeak2Thr=1200);

[0075] For a 3M card, select data points where Noise1 is greater than the threshold cNoise1Thr (e.g., cNoise1Thr=600) and Peak2 is greater than the threshold cPeak2Thr (e.g., cPeak2Thr=1200);

[0076] For the black card and white card fusion part, select data points where Noise1 is greater than the threshold bNoise1Thr (e.g., bNoise1Thr=500), less than wNoise1Thr (e.g., wNoise1Thr=1300), and Peak2 is less than the threshold cPeak2Thr (e.g., cPeak2Thr=1200).

[0077] S25. For each error correction segment, the ranging error is fitted using a preset lidar ranging correction model to obtain the error correction parameters corresponding to each error correction segment; wherein, the error correction parameters corresponding to each error correction segment together constitute the error correction parameter set.

[0078] The preset lidar ranging correction models include linear function models and quadratic function models, etc.

[0079] When the model is a linear function, Formula 1 is as follows:

[0080] ;

[0081] Noise1 is the noise value, DisErr is the ranging error, and k and b are the parameters to be solved by fitting the calibration data using the least squares method.

[0082] When the model is a quadratic function, Formula 2 is as follows:

[0083] ;

[0084] A, B, and C are the parameters to be solved obtained by fitting, which can also be obtained by fitting calibration data using the least squares method.

[0085] Each error correction segment has its own set of fitting parameters. If a linear function model is used, each segment has a set of k and b; if a quadratic function model is used, each segment has a set of A, B, and C. Therefore, the set of error correction parameters obtained throughout the calibration phase is actually a set of piecewise parameters. , where n is the total number of error correction segments.

[0086] Therefore, by fitting the data of each error correction segment separately, the correction parameters of the corresponding linear function model or quadratic function model can be obtained, thereby achieving accurate ranging error correction for targets with different noise levels and echo characteristics.

[0087] Next, in the application phase of the ranging method, please refer to... Figure 7 , Figure 7 This is a flowchart of a lidar ranging method provided in an embodiment of this application. Figure 7 As shown, the method includes:

[0088] S31. When the lidar is working, acquire ranging data; the ranging data includes noise value, echo characteristics and raw ranging results;

[0089] S32. When it is determined that the original ranging result needs to be corrected based on the noise value, the error correction segment to which the ranging data belongs is determined based on the echo characteristics, and the error correction parameter corresponding to the error correction segment is selected.

[0090] S33. Substitute the error correction parameters into the preset lidar ranging correction model to determine the corresponding ranging error correction function;

[0091] S34. Input the noise value into the ranging error correction function to calculate the distance error value;

[0092] S35. Correct the original ranging result based on the distance error value to obtain the target ranging result of the lidar.

[0093] The lidar includes a photosensitive chip, which comprises a first region and a second region. The first region is the normal receiving region, used to receive echo signals and output noise values. The second region is the weakened receiving region, used to attenuate the energy of the received echo signals and output echo peak values. When the lidar is operating, acquiring ranging data specifically includes: acquiring the noise value output from the first region; acquiring the echo peak value output from the second region; and acquiring the raw ranging result corresponding to the noise value and echo peak value.

[0094] Specifically, when it is determined that the original ranging result needs to be corrected based on the noise value, the error correction segment to which the ranging data belongs is determined based on the echo characteristics, and the error correction parameter corresponding to the error correction segment is selected. This includes: comparing the noise value with a preset noise threshold; if the noise value is greater than the preset noise threshold, it is determined that the original ranging result needs to be corrected; when it is determined that the original ranging result needs to be corrected, comparing the echo characteristics with at least one preset echo characteristic threshold, and dividing the ranging data into the corresponding error correction segment based on the comparison result; and selecting the error correction parameter corresponding to the error correction segment from the pre-calibrated set of error correction parameters according to the error correction segment to which the ranging data belongs.

[0095] The lidar ranging correction model can be either a linear function model or a quadratic function model. Each error correction segment has its model used for fitting predetermined during the calibration phase. For example, some segments might use linear fitting to obtain k and b during calibration, and those segments would then use a linear function model in the application phase. Therefore, the predetermined model and parameters corresponding to the current ranging data's error correction segment can be directly used.

[0096] Specifically, the process of correcting the original ranging result based on the distance error value to obtain the target ranging result of the lidar includes: performing a subtraction operation between the original ranging result and the distance error value, and using the difference as the target ranging result of the lidar.

[0097] For example, by choosing a linear function model, we can obtain the fitting parameters wP2_n1 for the white card segment, such as wP2_n1=[0.0106-28.6241], the fitting parameters cP2_n1 for the 3M card segment, such as cP2_n1=[0.0222-68.9685], and the fitting parameters wbP2_n1 for the fused white and black card segment, such as wbP2_n1=[-0.0021-5.4149]. Substituting these fitting parameters into the linear function model, i.e., Formula 1 above, we can obtain the distance error values ​​for different segments.

[0098] For example, white card segmentation: distance error value = Noise1 * 0.0106 + (-28.6241);

[0099] 3M card segmentation: Distance error value = Noise1 * 0.0222 + (-68.9685);

[0100] White card and black card fusion segment: distance error value = Noise1*(-0.0021)+(-5.4149).

[0101] The target ranging result of lidar is obtained by subtracting the original ranging result corresponding to each segment from the distance error value.

[0102] Therefore, the above-mentioned lidar ranging method can make ranging unaffected by strong light during sensor operation, thus improving the ranging accuracy and stability of lidar under complex lighting conditions.

[0103] It should be noted that in the above embodiments, there is no necessarily a certain order between the steps. Those skilled in the art can understand from the description of the embodiments of this application that the above steps may have different execution orders in different embodiments, that is, they may be executed in parallel or in turn, etc.

[0104] Please see Figure 8 , Figure 8 This is a schematic diagram of a lidar system provided in an embodiment of this application. The lidar 12 includes one or more processors 121 and a memory 122. The memory 122 is connected to one or more processors 121, for example, via a bus.

[0105] Processor 121 is configured to support the lidar 12 in performing the corresponding functions in the methods described in the above-described method embodiments. Processor 121 may be a central processing unit (CPU), a network processor (NP), a hardware chip, or any combination thereof. The aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The aforementioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0106] Memory 122 is used to store program code, etc. Memory 122 may include volatile memory (VM), such as random access memory (RAM); memory may also include non-volatile memory (NVM), such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); memory 122 may also include combinations of the above types of memory.

[0107] The memory 122 can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the lidar ranging method in the embodiments of this application. The processor 121 executes various functional applications and data processing of the lidar ranging method by running the non-volatile software programs, instructions, and modules stored in the memory 122, thereby implementing the lidar ranging method provided in the above-described method embodiments.

[0108] The memory 122 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and application programs required for at least one function. One or more modules are stored in the memory 122, and when executed by one or more processors 121, they execute the lidar ranging method in any of the above method embodiments, for example, executing... Figure 7 The methods and steps described.

[0109] In some embodiments, the lidar 12 described above can be replaced by a ranging correction device. This ranging correction device includes, but is not limited to, a signal processing module integrated within the lidar 12, a lidar main control chip, a microcontroller unit, a digital signal processor, a system-on-a-chip, a lidar control board, and a lidar sensor module. Alternatively, it can be an external processing unit, edge computing device, vehicle controller, robot controller, industrial controller, or server that is communicatively connected to the lidar 12.

[0110] In one embodiment, the lidar 12 can directly execute the aforementioned ranging method without requiring additional devices. In this case, the ranging correction function inside the lidar 12 can be implemented by an internal signal processing module, lidar main control chip, microcontroller unit (MCU), digital signal processor (DSP), system-on-a-chip (SoC), lidar control board, or lidar sensor module, etc., to complete noise detection, echo feature analysis, error correction segment selection, and ranging correction calculation, thereby realizing real-time ranging error correction of the lidar.

[0111] In another embodiment, the ranging correction function can be performed by an external processing unit communicatively connected to the lidar 12. This external processing unit may include edge computing devices, vehicle controllers, robot controllers, industrial controllers, servers, etc. The lidar sends real-time ranging data (including noise values, echo characteristics, and the original ranging result) to the external processing unit, which then executes a ranging error correction model, calculates the corrected target ranging result, and returns the corrected result to the lidar or other downstream systems, thereby improving ranging accuracy under complex lighting conditions.

[0112] The data calibration stage can be applied to the range correction device integrated inside the lidar 12, or it can be an error correction parameter set provided by an external processing unit.

[0113] This application provides a non-volatile computer-readable storage medium storing computer-executable instructions. These computer-executable instructions are then accessed by a lidar (e.g., a laser radar). Figure 8 When the processor 121 shown is executed, it causes the lidar to perform the lidar ranging method as described above; when the computer-executable instruction is executed by the ranging correction device, it causes the ranging correction device to perform the lidar ranging method as described above.

[0114] This application provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium. The computer program includes program instructions, which, when executed by the lidar or the ranging correction device, can realize the lidar ranging method as described above.

[0115] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0116] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the claims of this application shall still fall within the scope of this application.

Claims

1. A lidar ranging method, characterized in that, include: When the lidar is working, it acquires ranging data; the ranging data includes noise value, echo characteristics, and raw ranging results. The noise value is compared with a preset noise threshold. If the noise value is greater than the preset noise threshold, it is determined that the original ranging result needs to be corrected. The preset noise threshold is determined during the calibration stage based on the statistical distribution relationship between the noise value and the ranging error. If it is determined that the original ranging result needs to be corrected, the echo feature is compared with at least one preset echo feature threshold, and the ranging data is divided into corresponding error correction segments according to the comparison result; wherein, the preset echo feature threshold is determined during the calibration stage based on the statistical distribution relationship between the echo feature and the ranging error; Based on the error correction segment to which the ranging data belongs, select the error correction parameter corresponding to the error correction segment from the pre-calibrated set of error correction parameters; Substitute the error correction parameters into the preset lidar ranging correction model to determine the corresponding ranging error correction function; The noise value is input into the ranging error correction function to calculate the distance error value. The original ranging result is corrected based on the distance error value to obtain the target ranging result of the lidar.

2. The method according to claim 1, characterized in that, The lidar includes a photosensitive chip, which includes a first region and a second region. The first region is a normal receiving region, used to receive echo signals and output noise values. The second region is a weakened receiving region, used to perform energy attenuation processing on the received echo signals and output echo peak values. The process of acquiring ranging data during lidar operation includes: Obtain the noise value output from the first region; Obtain the peak echo output from the second region; Obtain the original ranging results corresponding to the noise value and the echo peak value.

3. The method according to claim 1, characterized in that, The step of correcting the original ranging result based on the distance error value to obtain the target ranging result of the lidar includes: The difference between the original ranging result and the distance error value is used as the target ranging result of the lidar.

4. The method according to claim 1, characterized in that, The method further includes: obtaining the error correction parameter set; The step of obtaining the error correction parameter set includes: Acquire multiple sets of calibration ranging data, including noise values, echo characteristics, original ranging results, and true distances; The ranging error is calculated based on the original ranging result and the actual distance. Based on the noise value and the echo characteristics, the multiple sets of calibration ranging data are segmented to obtain multiple error correction segments; Extract the ranging error of each of the aforementioned error correction segments; For each error correction segment, a preset lidar ranging correction model is used to fit the ranging error to obtain the error correction parameters corresponding to each error correction segment; wherein, the error correction parameters corresponding to each error correction segment together constitute the error correction parameter set.

5. The method according to claim 4, characterized in that, The lidar is mounted on a calibration fixture, and a calibration target is positioned at a preset distance in front of the lidar. The calibration target includes targets with different reflection characteristics, including at least low-reflection targets, medium-reflection targets, and high-reflection targets. A light source is positioned within a preset range of the calibration target, and the light source is used to illuminate the calibration target with different light intensities. The acquisition of multiple sets of calibration ranging data includes: The actual distance between the lidar and the calibrated target is controlled, and the noise value, echo characteristics, and original ranging results are collected at multiple different locations with different actual distances, as well as under different target objects and different light intensity conditions; wherein, the noise value, echo characteristics, and original ranging results obtained in the same ranging event collectively correspond to the actual distance, and the actual distance is taken as the true distance.

6. The method according to claim 4, characterized in that, The process of segmenting the multiple sets of calibration ranging data based on the noise value and the echo characteristics yields multiple error correction segments, including: By statistically analyzing the correspondence between the noise value and the ranging error in the multiple sets of calibration ranging data, the distribution of the noise value and the ranging error is obtained; Based on the distribution of the noise value and the ranging error, at least one noise threshold is determined, and the multiple sets of calibration ranging data are initially segmented according to the noise threshold. After completing the initial segmentation, the correspondence between the echo features and the ranging error in each initial segment is analyzed to obtain the distribution of the echo features and the ranging error. Based on the distribution of the echo characteristics and the ranging error, at least one echo characteristic threshold is determined, and the preliminary segment is further subdivided based on the echo characteristic threshold to obtain multiple error correction segments.

7. A lidar, characterized in that, include: A memory and a processor, the memory being connected to the processor, the processor being configured to execute one or more computer programs stored in the memory, the processor, when executing the one or more computer programs, causing the lidar to perform the method as described in any one of claims 1-3.

8. A ranging correction device, characterized in that, include: A memory and a processor, wherein the memory is connected to the processor. The processor is configured to receive ranging data output by the lidar and execute one or more computer programs stored in the memory based on the ranging data. When the processor executes the one or more computer programs, it causes the ranging correction device to implement the method as described in any one of claims 1-6.

9. A non-volatile computer-readable storage medium, characterized in that, The non-volatile computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are executed by the lidar, the lidar performs the method according to any one of claims 1-3; When the computer-executable instructions are executed by the ranging correction device, the ranging correction device performs the method according to any one of claims 1-6.

Images