Laser ranging method, device and equipment

Abstract

The invention provides a laser ranging method, device and equipment, and the method comprises the steps: obtaining a histogram based on the detection data of a laser radar, the horizontal coordinate of the histogram being the flight time of a laser pulse, and the vertical coordinate of the histogram being the frequency corresponding to the flight time; acquiring shape features corresponding to peak points of the histogram, wherein the shape features comprise peak height, left peak width and right peak width; wherein the peak height is the difference value between the height of the peak highest point and the height of the histogram background, the left peak width is the width of the position from the position of the peak highest point to the position after the height of the left side of the peak is reduced by a first height, and the right peak width is the width of the position from the position of the peak highest point to the position after the height of the right side of the peak is reduced by a second height; determining the peak point as an object peak or an interference peak based on the shape features; and if the peak point is the object peak, determining the object distance based on the flight time corresponding to the peak point. According to the technical scheme of the invention, false alarm caused by a rain and fog interference peak when the laser radar works in a rain and fog scene is effectively avoided, and finally the purpose of improving the distance measurement accuracy is achieved.

Description

Technical Field

[0001] The present application relates to the field of laser ranging technology, and in particular to a laser ranging method, device and equipment. Background Art

[0002] LiDAR is a non-contact distance measurement device. Its principle is as follows: the LiDAR emits a laser pulse, which strikes an object and is reflected by it. After reflection, the laser pulse is detected again by the LiDAR. The LiDAR measures the time of flight (t) between the laser pulse's emission and reception. Based on this time of flight, the distance to the object is determined as d = ct / 2, where c represents the speed of light.

[0003] However, for outdoor rain and fog scenes (such as rainfall scenes or water mist scenes), during the laser ranging process, the laser pulses emitted by the lidar will be scattered when they hit raindrops or water mist, and some backscattered light will be received by the lidar, causing the lidar to mistakenly believe that there are objects in the raindrops or water mist, resulting in false alarms, and the ranging function cannot be realized normally, and thus accurate ranging results cannot be obtained. Summary of the Invention

[0004] The present application provides a laser ranging method, which is applied to a laser radar. The method includes:

[0005] A histogram is obtained based on the detection data of the laser radar, wherein the abscissa of the histogram is the flight time of the laser pulse, and the ordinate of the histogram is the frequency corresponding to the flight time;

[0006] Obtaining shape features corresponding to the peak points of the histogram, the shape features including peak height, left peak width, and right peak width; wherein the peak height is the difference between the height of the highest point of the peak and the background height of the histogram, the left peak width is the width from the position of the highest point of the peak to the position where the height on the left side of the peak drops by a first height, and the right peak width is the width from the position of the highest point of the peak to the position where the height on the right side of the peak drops by a second height;

[0007] determining, based on the shape feature, that the peak point is an object peak or an interference peak;

[0008] If the peak point is an object peak, the object distance is determined based on the flight time corresponding to the peak point.

[0009] The present application provides a laser ranging device for use in a laser radar, the device comprising:

[0010] an acquisition module, configured to acquire a histogram based on the detection data of the laser radar, wherein the abscissa of the histogram is the flight time of the laser pulse, and the ordinate of the histogram is the frequency corresponding to the flight time; and to acquire shape features corresponding to the peak points of the histogram, wherein the shape features include peak height, left peak width, and right peak width; wherein the peak height is the difference between the height of the highest point of the peak and the height of the background of the histogram, the left peak width is the width from the position of the highest point of the peak to the position where the height on the left side of the peak drops by a first height, and the right peak width is the width from the position of the highest point of the peak to the position where the height on the right side of the peak drops by a second height;

[0011] A determination module is configured to determine whether the peak point is an object peak or an interference peak based on the shape feature; if the peak point is an object peak, determine the object distance based on the flight time corresponding to the peak point.

[0012] The present application provides an electronic device, comprising: a processor and a machine-readable storage medium, wherein the machine-readable storage medium stores machine-executable instructions that can be executed by the processor; the processor is used to execute the machine-executable instructions to implement the laser ranging method of the above example of the present application.

[0013] As can be seen from the above technical solution, in the embodiment of the present application, by extracting the shape features corresponding to the peak points of the histogram (such as peak height, left peak width and right peak width), the peak points are determined to be object peaks or interference peaks based on the shape features, and the peak points can be classified, thereby effectively avoiding false alarms caused by rain and fog interference peaks when the laser radar is working in rain and fog scenes. The laser radar has the ability to classify the detected peaks, reduce the false alarms of the laser radar in rain and fog environments, filter out interference from rain and fog, and ultimately achieve the purpose of improving the ranging accuracy. The ranging function can be realized normally, so that accurate ranging results can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments of the present application or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings of the embodiments of the present application.

[0015] Figure 1 This is a flow chart of a laser ranging method in one embodiment of the present application;

[0016] Figure 2 This is a flow chart of a laser ranging method in one embodiment of the present application;

[0017] Figure 3 This is a schematic diagram of the ranging principle and histogram of a laser radar in one embodiment of the present application;

[0018] Figure 4 This is a schematic diagram of the ranging principle and histogram of a laser radar in one embodiment of the present application;

[0019] Figure 5 is a schematic diagram of shape features corresponding to peak points of a histogram in one embodiment of the present application;

[0020] Figure 6 is a schematic diagram of a classification problem of points in a feature space in one embodiment of the present application;

[0021] Figure 7 1 is a schematic structural diagram of a laser ranging device in one embodiment of the present application;

[0022] Figure 8 It is a hardware structure diagram of an electronic device in one embodiment of the present application. DETAILED DESCRIPTION

[0023] The terms used in the embodiments of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms "a," "the," and "the" used in this application and claims are also intended to include plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein refers to any or all possible combinations of one or more associated listed items.

[0024] It should be understood that although the terms first, second, third, etc. may be used to describe various information in the embodiments of the present application, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" used may also be interpreted as "at the time of" or "when" or "in response to determining".

[0025] In the embodiment of the present application, a laser ranging method is proposed, which can be applied to laser radar. Figure 1 FIG. 1 is a flow chart of the laser ranging method, which may include:

[0026] Step 101: Obtain a histogram based on the detection data of the laser radar. The horizontal axis of the histogram can be the flight time of the laser pulse, and the vertical axis of the histogram can be the frequency corresponding to the flight time.

[0027] Step 102: Obtain shape features corresponding to the peaks of the histogram. The shape features may include peak height, left peak width, and right peak width. The peak height is the difference between the height of the highest point of the peak and the background height of the histogram. The left peak width is the width from the highest point of the peak to the position where the height on the left side of the peak drops by a first height. The right peak width is the width from the highest point of the peak to the position where the height on the right side of the peak drops by a second height.

[0028] Step 103: Determine whether the peak point is an object peak or an interference peak based on the shape feature.

[0029] Step 104: If the peak point is an object peak, determine the object distance based on the flight time corresponding to the peak point.

[0030] Exemplarily, determining whether the peak point is an object peak or an interference peak based on the shape feature may include, but is not limited to: determining a first eigenvalue based on the left peak width, determining a second eigenvalue based on the right peak width, and determining a third eigenvalue based on the peak height; wherein the first eigenvalue and the second eigenvalue may both be greater than or equal to 0, and the third eigenvalue may be less than 0; wherein, the larger the left peak width, the larger the first eigenvalue; the larger the right peak width, the larger the second eigenvalue; and the larger the peak height, the smaller the absolute value of the third eigenvalue. A target eigenvalue is determined based on the first eigenvalue, the second eigenvalue, and the third eigenvalue, and the peak point is determined to be an object peak or an interference peak based on the target eigenvalue.

[0031] Exemplarily, the calibrated support group may include N parameter groups, where N may be a positive integer, and each parameter group may include a left peak width coefficient, a right peak width coefficient, and a mapping table, and the mapping table may include a correspondence between a calibrated peak height and a peak height characteristic value. Determining whether the peak point is an object peak or an interference peak based on the shape feature may include, but is not limited to: for each parameter group, determining a first characteristic value based on the product value of the left peak width and the left peak width coefficient, determining a second characteristic value based on the product value of the right peak width and the right peak width coefficient, and obtaining a third characteristic value corresponding to the peak height by querying the mapping table through the peak height. Determine the target characteristic value corresponding to the parameter group based on the first characteristic value, the second characteristic value, and the third characteristic value. If the target characteristic value corresponding to each parameter group is less than a threshold value (such as 0), the peak point is determined to be an object peak. If the target characteristic value corresponding to any parameter group is not less than a threshold value, the peak point is determined to be an interference peak.

[0032] Exemplarily, for each parameter group within a support group, the calibration process of the parameter group may include, but is not limited to: using the configured typical value of the left peak width coefficient as the left peak width coefficient in the parameter group, and using the configured typical value of the right peak width coefficient as the right peak width coefficient in the parameter group. Obtaining a sample data set, the sample data set may include a first shape feature corresponding to the sample object peak and a second shape feature corresponding to the sample interference peak; for each calibration peak height, selecting the first shape feature and the second shape feature corresponding to the calibration peak height from the sample data set. Based on the left peak width coefficient, the right peak width coefficient, the first shape feature corresponding to the calibration peak height, and the second shape feature corresponding to the calibration peak height, determine the peak height characteristic value corresponding to the calibration peak height, and record the correspondence between the calibration peak height and the peak height characteristic value in the mapping table of the parameter group.

[0033] Illustratively, determining the peak height characteristic value corresponding to the calibrated peak height based on the left peak width coefficient, the right peak width coefficient, the first shape feature corresponding to the calibrated peak height, and the second shape feature corresponding to the calibrated peak height may include, but is not limited to: if the feature space region of the first shape feature corresponding to the sample object peak does not overlap with the feature space region of the second shape feature corresponding to the sample interference peak, determining the peak height characteristic value using the following expression: Alternatively, if the feature space region of the first shape feature corresponding to the sample object peak coincides with the feature space region of the second shape feature corresponding to the sample interference peak, the peak height feature value can be determined using the following expression: Among them, α i represents the left peak width coefficient, β i represents the right peak width coefficient, They represent the left peak width, right peak width and peak height corresponding to the j-th sample object peak, respectively. They represent the left peak width, right peak width and peak height corresponding to the j-th sample interference peak; I′ represents the calibration peak height, γ i (I′) represents the peak height characteristic value corresponding to the calibrated peak height.

[0034] Exemplarily, M statistical results corresponding to the peak point can also be obtained based on M consecutive histograms, where M can be a positive integer. For each statistical result, the statistical result can indicate whether the peak point is an object peak or an interference peak. The number of occurrences of the object peak can be determined based on the M statistical results. If the ratio of the number of occurrences of the object peak to M is greater than or equal to a first threshold (configured based on experience), the peak point can be determined to be an object peak. If the ratio of the number of occurrences of the object peak to M is less than the first threshold, the peak point can be determined to be an interference peak. Alternatively, the number of occurrences of the interference peak can be determined based on the M statistical results. If the ratio of the number of occurrences of the interference peak to M is less than or equal to a second threshold (configured based on experience), the peak point can be determined to be an object peak. If the ratio of the number of occurrences of the interference peak to M is greater than the second threshold, the peak point can be determined to be an interference peak.

[0035] Exemplarily, the laser ranging method can be used to determine the distance of an object in a target scene, and the target scene may include but is not limited to a rain and fog scene; wherein the interference peak is an interference peak generated by rain and / or fog.

[0036] As can be seen from the above technical solution, in the embodiment of the present application, by extracting the shape features corresponding to the peak points of the histogram (such as peak height, left peak width and right peak width), the peak points are determined to be object peaks or interference peaks based on the shape features, and the peak points can be classified, thereby effectively avoiding false alarms caused by rain and fog interference peaks when the laser radar is working in rain and fog scenes. The laser radar has the ability to classify the detected peaks, reduce the false alarms of the laser radar in rain and fog environments, filter out interference from rain and fog, and ultimately achieve the purpose of improving the ranging accuracy. The ranging function can be realized normally, so that accurate ranging results can be obtained.

[0037] The laser ranging method of the embodiment of the present application is described below in conjunction with specific application scenarios.

[0038] The ranging principle of LiDAR is as follows: the LiDAR emits a laser pulse, which is reflected by the object after hitting it. After reflection, the laser pulse is detected again by the LiDAR. The LiDAR measures the flight time t from the laser pulse's emission to its reception, and based on the flight time t, determines the distance to the object as d = ct / 2, where c represents the speed of light. However, in outdoor rainy and foggy scenes, during the laser ranging process, the laser pulse emitted by the LiDAR will scatter when it hits raindrops or water mist. Some of the backscattered light will be received by the LiDAR, causing the LiDAR to mistakenly believe that there is an object in the raindrops or water mist, resulting in a false alarm.

[0039] In response to the above findings, a laser ranging method is proposed in an embodiment of the present application, which can be applied to a laser radar. The laser radar can be a single-point laser radar (that is, the detection module of the laser radar is composed of only a single pixel and has no spatial resolution capability) or a multi-point laser radar, without any restriction.

[0040] Exemplarily, the laser ranging method can be used to determine the distance to an object in a target scenario, and the target scenario may include, but is not limited to, a rainy or foggy scenario. A rainy or foggy scenario can be an application scenario in which rain is present, or a rainy or foggy scenario can be an application scenario in which mist is present, or a rainy or foggy scenario can be an application scenario in which both rain and mist are present. Of course, the above are just a few examples of target scenarios, and there is no limitation to these target scenarios. The target scenario can be any application scenario, and the laser ranging method can be implemented in any application scenario.

[0041] In the embodiment of the present application, a laser ranging method is proposed. Figure 2 As shown, the method may include:

[0042] Step 201: Obtain a histogram based on the detection data of the laser radar. The horizontal axis of the histogram can be the flight time of the laser pulse, and the vertical axis of the histogram can be the frequency corresponding to the flight time.

[0043] For example, see Figure 3 As shown in the figure, the left side is a schematic diagram of the ranging principle of the laser radar. The laser radar (such as a single-point laser radar) can emit a laser pulse, which is reflected by the object. After reflection, the laser pulse is detected again by the laser radar. The laser radar measures the flight time t of the laser pulse from emission to reception. On this basis, the distance to the object is d = ct / 2, and c is the speed of light.

[0044] In order to improve the detection signal-to-noise ratio, the laser radar will emit multiple laser pulses during each frame of detection. The flight time of multiple laser pulses can be recorded, and a histogram can be obtained based on the flight time results. The horizontal axis of the histogram can be the flight time of the laser pulse (i.e., flight time t), and the vertical axis of the histogram can be the frequency corresponding to the flight time (the frequency indicates how many laser pulses correspond to the flight time).

[0045] Exemplarily, the detection data of the laser radar can be the flight time of multiple laser pulses, that is, after emitting multiple laser pulses, the flight time of the multiple laser pulses is used as the detection data of the laser radar. Of course, the detection data can also include other types of data, and there is no limitation on this detection data.

[0046] See also Figure 3As shown in the figure, the right side is a schematic diagram of the histogram. For each flight time, the flight time corresponds to a frequency. For example, assuming that the lidar emits 100 laser pulses, the flight time of 50 laser pulses is flight time A, the flight time of 30 laser pulses is flight time B, and the flight time of 20 laser pulses is flight time C. When obtaining the histogram, the frequency corresponding to flight time A is 50, the frequency corresponding to flight time B is 30, and the frequency corresponding to flight time C is 20.

[0047] A histogram is a statistical graph generated by multiple timestamp data, showing the number of times a lidar is triggered in different time periods. The timestamp is the time data recorded by the lidar when it is triggered by the echo signal.

[0048] See also Figure 3 As shown in the figure, assuming there is one object in the LiDAR scene, the object will form an object peak on the histogram. That is, the frequency of the object peak is greater than the frequency to the left of the object peak, and the frequency of the object peak is greater than the frequency to the right of the object peak. Assuming there are at least two objects in the LiDAR scene, each object will form an object peak on the histogram. That is, each object corresponds to an object peak.

[0049] In summary, each peak on the histogram corresponds to an object to be measured in the scene to be measured, so the distance of the object to be measured is determined based on the flight time corresponding to the peak, such as the object distance d=ct / 2.

[0050] However, in outdoor applications, LiDAR often needs to operate in rainy or foggy conditions (such as rainfall or mist). During laser ranging, when the laser pulses emitted by the LiDAR hit raindrops or mist, they scatter. Some of the backscattered light is received by the LiDAR, causing it to mistakenly identify an object as being in the raindrops or mist, leading to false alarms.

[0051] See also Figure 4 As shown in the figure, the left side is a schematic diagram of the ranging principle of the laser radar. The laser radar (such as a single-point laser radar) can emit laser pulses. When the laser pulses hit raindrops or water mist, they will be scattered. Part of the backscattered light will be received by the laser radar, causing the laser radar to mistakenly believe that there are objects in the raindrops or water mist.

[0052] See also Figure 4 As shown in the figure, the right side is a schematic diagram of the histogram. When the laser pulse is irradiated by raindrops or water mist, it will be scattered. After the detection results of multiple laser pulses are accumulated, a rain and fog interference peak is formed in the histogram. The rain and fog interference peak is mistaken by the lidar as the presence of an object here, resulting in a false alarm.

[0053] In view of the above findings, in this embodiment, subsequent steps can also be used to distinguish object peaks from interference peaks (such as rain and fog interference peaks), so that the location of the object peak is identified as an object, and the location of the interference peak is not identified as an object, thereby removing false alarms caused by rain and fog interference peaks and avoiding erroneous detection results.

[0054] Step 202: Obtain the peak point of the histogram, where the peak point is the position of the peak of the histogram. For example, the peak point may be the position of the object peak of the histogram or the position of the interference peak of the histogram.

[0055] Exemplarily, after obtaining the histogram, a peak point can be obtained from the histogram, where the frequency of the peak point is greater than the frequency to the left of the peak point, and the frequency of the peak point is greater than the frequency to the right of the peak point.

[0056] The number of peak points may be one or more. Since the processing process for each peak point is the same, for the convenience of description, the processing process of one peak point is taken as an example.

[0057] Step 203: Obtain shape features corresponding to the peaks of the histogram. The shape features may include peak height, left peak width, and right peak width. The peak height is the difference between the height of the highest point of the peak and the background height of the histogram. The left peak width is the width from the highest point of the peak to the position where the height on the left side of the peak drops by a first height. The right peak width is the width from the highest point of the peak to the position where the height on the right side of the peak drops by a second height.

[0058] In order to distinguish object peaks from interference peaks, we first analyze the characteristics of object peaks and interference peaks:

[0059] 1. Real objects can be regarded as "hard targets", that is, the reflection of laser pulses by real objects only occurs on the surface of real objects. The measurement results of the flight time of multiple laser pulses have small deviations from each other, and the peak formed on the histogram (i.e., the object peak) is narrower, and its width is roughly equivalent to the pulse width of the laser pulse.

[0060] 2. Scattering environments such as rain and fog can be considered "soft targets," typically occupying a large spatial area. When a laser pulse enters a scattering environment such as rain and fog, it will be scattered once or multiple times at different locations within this scattering environment. As a result, the flight times received by the lidar vary significantly, resulting in wider peaks (i.e., interference peaks) on the histogram. Their width is roughly related to the mean free path of the scattering environment. The mean free path is defined as the average distance a laser pulse travels between two scatterings in the scattering environment.

[0061] 3. The difference in peak width of object peaks or interference peaks (such as rain and fog interference peaks) also contains a certain asymmetry, that is, the width of the rising edge and the width of the falling edge of the object peak or interference peak also have a certain difference.

[0062] The reason for this phenomenon is that the laser pulses emitted by the lidar are inherently asymmetric, that is, the width of the rising edge and the width of the falling edge of the laser pulse are different (usually due to the charging and discharging characteristics of the laser pump current). It is caused by the pile-up effect of the lidar. For example, when the lidar uses a single-photon avalanche diode (SPAD) or a silicon photomultiplier (SiPM), the lidar will experience the so-called "dead time" when converting the received laser pulse to flight time. That is, after undergoing a photon-electron conversion process, there will be a period of time when it cannot respond to subsequent photons. This causes the laser pulses obtained by multiple measurements to have a pile-up effect, that is, the probability of measuring the rising edge of the pulse is higher, resulting in the measured histogram peak with a steeper rising edge and a flatter falling edge, further exacerbating this asymmetry.

[0063] Based on the above analysis, it can be seen that the shape of the object peak (represented by the shape feature) is different from the shape of the interference peak (represented by the shape feature). Therefore, the object peak and the interference peak can be distinguished by their different shapes. For example, the shape feature corresponding to the peak point of the histogram can be obtained, and then the shape feature corresponding to the peak point can be used to distinguish whether the peak point is an object peak or an interference peak.

[0064] For example, for a peak point in the histogram, the shape feature corresponding to the peak point can be represented by peak height, left peak width, and right peak width. That is, the shape feature can include peak height, left peak width, and right peak width. Of course, peak height, left peak width, and right peak width are only a few examples of shape features and are not limited to these shape features.

[0065] See also Figure 5 Figure 1 shows a schematic diagram of the shape features corresponding to histogram peaks. Peak height I is the difference between the peak's highest point and the histogram background height. Peak height I is defined as the peak's highest point minus the histogram background height. For example, the peak's highest point height can be the frequency corresponding to the histogram peak. The histogram background height can be a pre-configured frequency representing frequencies generated by interference such as noise and ambient light.

[0066] For example, when there are no objects in the scene, the lidar emits multiple laser pulses to obtain a histogram of the scene. The average of the frequencies corresponding to all peaks in the histogram is used as the histogram background height. In other words, the histogram background height represents the average of the frequencies when no objects are present. Of course, the above is only an example of obtaining the histogram background height, and there is no limitation on the method of obtaining the histogram background height.

[0067] See also Figure 5 As shown, the left peak width w l It is the first height h from the highest point of the peak to the left side of the peak l The width of the rear position, that is, the left peak width wl It is defined as the height drop from the highest point of the peak to the left side of the peak h l For example, h l It can be configured based on experience. l There is no restriction on the value of h. l The value can be set from 0 to 1 according to the actual situation. l The value of is 0, then the left peak width w l is 0, if h l The value of is I, then the left peak width w l Indicates the width from the highest point of the peak to the frequency 0.

[0068] See also Figure 5 As shown, the right peak width w r It is the second height h from the highest point of the peak to the right side of the peak. r The width of the rear position, that is, the right peak width w r It is defined as the height drop from the highest point of the peak to the right side of the peak h r For example, h r It can be configured based on experience. r There is no restriction on the value of h. r The value can be set from 0 to 1 according to the actual situation. r The value of is 0, then the right peak width w r is 0, if h r The value of is I, then the right peak width w r Indicates the width from the highest point of the peak to the frequency 0.

[0069] For example, if the pile-up effect of the laser radar is serious, the rising edge of the histogram peak will be very steep and the falling edge will be very gentle. In this case, h l is I, then the left peak width is the entire rising edge width, which can be taken as h r =1 / 2, at this time the right peak width is about half of the falling edge width. In this way, the left and right peak widths are less affected by the histogram jitter, which can make the subsequent algorithm more robust. Of course, the above is only the first height h l and the second height h r For example, the first height h l and the second height h r There are no restrictions.

[0070] In one possible implementation, after obtaining the histogram, the shape features corresponding to the peaks of the histogram can be directly obtained. Alternatively, to reduce the impact of noise, the histogram can be smoothed and filtered, and after smoothing, the shape features corresponding to the peaks of the smoothed filtered histogram can be obtained.

[0071] When smoothing the histogram, for the frequency corresponding to each flight time of the histogram, the frequency after filtering (such as the average of these frequencies, or weighted calculation of these frequencies) can be determined based on the frequency of adjacent flight times (such as k adjacent flight times on the left, or k adjacent flight times on the right, or k1 adjacent flight times on the left and k2 adjacent flight times on the right) and the frequency of the flight time, and the frequency after filtering is used to replace the frequency corresponding to the flight time. Of course, the above is only an example of smoothing the histogram, and the implementation method of this smoothing filtering is not limited in this embodiment.

[0072] Step 204: Determine whether the peak point is an object peak or an interference peak based on the shape feature.

[0073] Exemplarily, the shape feature may include peak height, left peak width and right peak width. Based on the left peak width, a first eigenvalue (i.e., the eigenvalue of the left peak width) is determined. The first eigenvalue may be greater than or equal to 0, i.e., the first eigenvalue is a positive number or 0. Based on the right peak width, a second eigenvalue (i.e., the eigenvalue of the right peak width) is determined. The second eigenvalue may be greater than or equal to 0, i.e., the second eigenvalue is a positive number or 0. A third eigenvalue (i.e., the eigenvalue of the peak height) may be determined based on the peak height. The third eigenvalue may be less than 0, i.e., the third eigenvalue is a negative number. For the first eigenvalue, if the left peak width is larger, the first eigenvalue is larger. For the second eigenvalue, if the right peak width is larger, the second eigenvalue is larger. For the third eigenvalue, if the peak height is larger, the absolute value of the third eigenvalue is smaller.

[0074] After obtaining the first eigenvalue, the second eigenvalue and the third eigenvalue, the target eigenvalue can be determined based on the first eigenvalue, the second eigenvalue and the third eigenvalue. For example, the sum of the first eigenvalue, the second eigenvalue and the third eigenvalue can be used as the target eigenvalue.

[0075] After obtaining the target eigenvalue, the peak point can be determined as an object peak or an interference peak based on the target eigenvalue. For example, if the target eigenvalue is less than a threshold value (such as 0), the peak point is determined to be an object peak. If the target eigenvalue is not less than the threshold value (such as 0), the peak point is determined to be an interference peak.

[0076] For example, a support group can be pre-calibrated, and the calibrated support group can include N parameter groups, where N can be a positive integer. For each parameter group, the parameter group can include a left peak width coefficient, a right peak width coefficient, and a mapping table, and the mapping table can include a correspondence between calibrated peak heights and peak height characteristic values.

[0077] For each parameter group, a first eigenvalue can be determined based on the product of the left peak width and the left peak width coefficient (i.e., the left peak width coefficient within the parameter group), such as the product of the left peak width and the left peak width coefficient as the first eigenvalue. For example, the left peak width coefficient can be a positive number or 0, such as a value between 0 and 1, and the first eigenvalue can be greater than or equal to 0, that is, the first eigenvalue is a positive number or 0. For example, for the first eigenvalue, the larger the left peak width, the larger the first eigenvalue.

[0078] For each parameter group, a second characteristic value can be determined based on the product of the right peak width and the right peak width coefficient (i.e., the right peak width coefficient within the parameter group), such as the product of the right peak width and the right peak width coefficient as the second characteristic value. For example, the right peak width coefficient can be a positive number or 0, such as a value between 0 and 1, and the second characteristic value can be greater than or equal to 0, that is, the second characteristic value is a positive number or 0. For example, for the second characteristic value, the larger the right peak width, the larger the second characteristic value.

[0079] For each parameter group, the mapping table (i.e., the mapping table within the parameter group) can be queried using the peak height to obtain the third eigenvalue corresponding to the peak height. For example, since the mapping table within the parameter group can include a correspondence between a calibrated peak height and a peak height eigenvalue, the mapping table can be queried using the peak height to obtain the peak height eigenvalue corresponding to the peak height, and the peak height eigenvalue can be used as the third eigenvalue. For example, the third eigenvalue can be less than 0, i.e., the third eigenvalue can be a negative number. For the third eigenvalue, the greater the peak height, the smaller the absolute value of the third eigenvalue can be.

[0080] When searching the mapping table based on the peak height, if the mapping table contains a calibrated peak height equal to the peak height, the peak height characteristic value corresponding to the calibrated peak height is found. If the mapping table does not contain a calibrated peak height equal to the peak height, the peak height characteristic value corresponding to the calibrated peak height with the smallest difference from the peak height is found.

[0081] For each parameter group, after obtaining the first eigenvalue, the second eigenvalue and the third eigenvalue corresponding to the parameter group, the target eigenvalue corresponding to the parameter group can be determined based on the first eigenvalue, the second eigenvalue and the third eigenvalue. For example, the sum of the first eigenvalue, the second eigenvalue and the third eigenvalue can be used as the target eigenvalue corresponding to the parameter group.

[0082] If the target eigenvalues corresponding to each parameter group are all less than a threshold value (such as 0), the peak point is determined to be an object peak. If the target eigenvalues corresponding to any parameter group are not less than the threshold value, the peak point is determined to be an interference peak.

[0083] In a possible implementation, the target characteristic value corresponding to the parameter group i may be determined using the following expression (1), where the value of i may range from 1 to N. Of course, expression (1) is only an example.

[0084] f i (w l ,w r ,I)=α i w l +β i w r +γ i (I) Expression (1)

[0085] In expression (1), f i (w l ,w r ,I) represents the target eigenvalue corresponding to the i-th parameter group, w l Indicates the left peak width, w r represents the right peak width, and I represents the peak height. i Represents the left peak width coefficient in the i-th parameter group, which can be a positive number or 0, such as a value between 0 and 1. i Represents the right peak width coefficient in the i-th parameter group, which can be a positive number or 0, such as a value between 0 and 1. i (I) represents the peak height characteristic value corresponding to the peak height I, that is, the peak height characteristic value γ corresponding to the peak height I is obtained by querying the mapping table in the i-th parameter group through the peak height I i (I). In expression (1), α i w l It can represent the first eigenvalue, β i w r It can be expressed as the second eigenvalue, γ i (I) can represent the third eigenvalue.

[0086] From expression (1), we can see that the target eigenvalue is a linear combination of the left peak width and the right peak width, and γ i (I) is a function of peak height (represented by a mapping table, i.e., the mapping table represents the relationship between peak height and peak height eigenvalue). The target eigenvalue reflects how the peak width characteristics of the object peak and the interference peak change with the change of peak height.

[0087] For example, if α = 1, β = 0, it means that the inequality distinguishes the object peak and the interference peak by the left peak width. If α = 1, β = 1, it means that the inequality distinguishes the object peak and the interference peak by the entire peak width. If α = 0, β = 1, it means that the inequality distinguishes the object peak and the interference peak by the right peak width.

[0088] For example, for the first parameter group, i is 1, f1(wl ,w r ,I)=α1w l +β1w r +γ1(I), α1 represents the left peak width coefficient in the first parameter group, β1 represents the right peak width coefficient in the first parameter group, and γ1(I) represents the peak height characteristic value obtained by querying the mapping table in the first parameter group through peak height I.

[0089] For the second parameter group, i is 2, f2(w l ,w r ,I)=α2w l +β2w r +γ2(I), α2 represents the left peak width coefficient in the second parameter group, β2 represents the right peak width coefficient in the second parameter group, γ2(I) represents the peak height characteristic value obtained by querying the mapping table in the second parameter group through peak height I, and so on.

[0090] In summary, for each parameter group, based on the left peak width coefficient, right peak width coefficient and mapping table in the parameter group, (w l ,w r ,I) into expression (1) to obtain the target eigenvalue corresponding to the parameter group. If the target eigenvalue corresponding to each parameter group is less than the threshold (such as 0), the peak point is determined to be an object peak. If the target eigenvalue corresponding to any parameter group is not less than the threshold (such as 0), the peak point is determined to be an interference peak.

[0091] In one possible embodiment, the left peak width w l , right peak width w r The three features such as peak height I can correspond to a three-dimensional feature space. Each peak on the histogram corresponds to a point in the feature space, and the coordinates are (w l ,w r ,I), so that the problem of "distinguishing whether a peak on the histogram is an object peak or an interference peak" can be converted into a classification problem of points in the feature space, see Figure 6 As shown in FIG, this is a schematic diagram of converting the problem of “distinguishing whether a peak on the histogram is an object peak or an interference peak” into a classification problem of points in the feature space.

[0092] For the classification problem in the feature space, it is necessary to predetermine the area occupied by the object peak in the feature space. This area can be expressed as f by N inequalities. i (w l ,w r ,I)<0, i=1,2,…,N, N can be a positive integer, and i can represent the inequality number. For example, if N inequalities If both of the inequalities hold, it means that the peak represented by the point is the object peak. Otherwise, if any of the inequalities If it is not true, it means that the peak represented by this point is an interference peak (such as a rain and fog interference peak).

[0093] This inequality group can be called a support group. The more inequalities N in the support group, the more accurate the description of the area occupied by the object peak and the better the classification effect. When the number N is larger, the amount of calculation is greater and more parameters need to be stored. In actual use, trade-offs need to be made and the value of N is reasonably configured.

[0094] The general form of the support group is f i (w l ,w r ,I)=α i w l +β i w r +γ i (I), which is the linear combination of the left peak width and the right peak width, and γ i (I) is a function of peak height, reflecting how the peak width characteristics of the object peak and the interference peak change with peak height. For example, if α = 1 and β = 0, the inequality distinguishes the object peak from the interference peak by the left peak width. If α = 1 and β = 1, the inequality distinguishes the object peak from the interference peak by the entire peak width. If α = 0 and β = 1, the inequality distinguishes the object peak from the interference peak by the right peak width.

[0095] After the parameters of the support group are determined in the above manner, for each peak detected on the histogram, it can be determined whether the peak belongs to an object peak or an interference peak in the above manner, thereby avoiding false alarms caused by interference peaks.

[0096] In summary, the support group can be pre-calibrated, and the calibrated support group can include N parameter groups. For each parameter group, the parameter group can include a left peak width coefficient, a right peak width coefficient, and a mapping table, and the mapping table can include the correspondence between the calibrated peak height and the peak height characteristic value. On this basis, the target characteristic value corresponding to each parameter group i can be determined using expression (1). If the target characteristic value corresponding to each parameter group is less than a threshold value (such as 0), then the peak point can be determined to be an object peak. If the target characteristic value corresponding to any parameter group is not less than a threshold value (such as 0), then the peak point can be determined to be an interference peak.

[0097] Step 205: If the peak point is an object peak, determine the object distance based on the flight time corresponding to the peak point. For example, the object distance is: d = ct / 2, where c represents the speed of light and t represents the flight time corresponding to the peak point.

[0098] If the peak point is an interference peak, which is an interference peak caused by rain and / or fog, it can be recorded as a rain and fog interference peak. In this case, the peak point will not be identified as an object, and the object distance will not be determined based on the flight time corresponding to the peak point.

[0099] In summary, the object peak and the interference peak can be distinguished, so the object distance is determined based only on the flight time corresponding to the object peak, rather than the flight time corresponding to the interference peak. This reduces the false alarms of the lidar in rainy and foggy environments, filters out environmental interference such as rain and fog, and ultimately achieves the goal of improving ranging accuracy. The ranging function can be implemented normally, thus obtaining accurate ranging results.

[0100] In one possible implementation, if the feature space occupied by the feature points of an object peak overlaps with the feature space occupied by the feature points of an interference peak, the single-frame judgment result may contain a certain amount of false positives, that is, the interference peak may be mistakenly identified as an object peak. Based on this, interference peaks can also be filtered out by using the judgment results of multiple frames in the time domain. That is, an object is only considered detected when M consecutive frames are judged as object peaks. The number of frames M here can be adjusted according to the actual application scenario to avoid false positives caused by interference peaks.

[0101] On this basis, in this embodiment, for the same peak point, M statistical results corresponding to the peak point can also be obtained based on M consecutive histograms, where M can be a positive integer. For each statistical result, the statistical result can indicate whether the peak point is an object peak or an interference peak. For details on how to obtain the statistical results corresponding to the peak point based on the histogram, please refer to steps 201 to 205, and will not be repeated here.

[0102] The number of occurrences of the object peak is determined based on M statistical results (for example, if 5 of the M statistical results indicate that the peak point is an object peak, then the number of occurrences of the object peak is 5). If the ratio of the number of occurrences of the object peak to M is greater than or equal to a first threshold value (which can be configured based on experience), then the peak point can be determined to be an object peak. If the ratio of the number of occurrences of the object peak to M is less than the first threshold value, then the peak point can be determined to be an interference peak. For example, the first threshold value can be 80%, 90%, 100%, etc., and there is no restriction on this. If the first threshold value is 100%, then when all statistical results indicate that the peak point is an object peak, then the peak point can be determined to be an object peak. If any statistical result indicates that the peak point is an interference peak, then the peak point can be determined to be an interference peak.

[0103] Alternatively, the number of occurrences of the interference peak is determined based on M statistical results (e.g., if 3 of the M statistical results indicate that the peak point is an interference peak, then the number of occurrences of the interference peak is 3). If the ratio of the number of occurrences of the interference peak to M is less than or equal to a second threshold value (which can be configured based on experience), then the peak point can be determined to be an object peak. If the ratio of the number of occurrences of the interference peak to M is greater than the second threshold value, then the peak point can be determined to be an interference peak. For example, the second threshold value can be 20%, 10%, 0%, etc., and there is no restriction on this. If the second threshold value is 0%, then when all statistical results do not indicate that the peak point is an interference peak, then the peak point can be determined to be an object peak. If any statistical result indicates that the peak point is an interference peak, then the peak point can be determined to be an interference peak.

[0104] In one possible implementation, to calibrate a support group, a sample dataset can be obtained. This sample dataset includes first shape features corresponding to sample object peaks and second shape features corresponding to sample interference peaks. Based on this, the support group can be calibrated based on the sample dataset. Specifically, each parameter group within the support group can be calibrated based on the sample dataset. The following describes the support group calibration process.

[0105] Step S11: Obtain a sample data set, which may include a first shape feature corresponding to a sample object peak and a second shape feature corresponding to a sample interference peak. For example, for ease of distinction, the object peak during the calibration process may be recorded as a sample object peak, and the shape feature corresponding to the sample object peak may be recorded as a first shape feature, where the first shape feature may include peak height, left peak width, and right peak width. The interference peak during the calibration process may be recorded as a sample interference peak, and the shape feature corresponding to the sample interference peak may be recorded as a second shape feature, where the second shape feature may include peak height, left peak width, and right peak width.

[0106] For example, for a scene containing an object, a histogram can be obtained for that scene, and sample object peaks can be obtained from the histogram. The first shape features corresponding to the sample object peaks can also be obtained. Obviously, performing the above processing on a large number of histograms can obtain the first shape features corresponding to a large number of sample object peaks. The first shape features corresponding to these sample object peaks can then be recorded in the sample dataset.

[0107] For example, for a scene with rain and fog, a histogram can be obtained for that scene, and the sample interference peaks can be obtained from the histogram. The second shape features corresponding to the sample interference peaks can also be obtained. Obviously, by performing the above processing on a large number of histograms, the second shape features corresponding to a large number of sample interference peaks can be obtained. The second shape features corresponding to these sample interference peaks can then be recorded in the sample dataset.

[0108] In one possible implementation, a large number of sample object peaks under different conditions (e.g., different object reflectivities, different distances, different rain and fog scattering intensities, etc.) can be calculated using a simulation method such as Monte Carlo or a probability of detection method, and first shape features corresponding to these sample object peaks can be obtained and recorded in a sample data set. Furthermore, a large number of sample interference peaks under different conditions (e.g., different object reflectivities, different distances, different rain and fog scattering intensities, etc.) can be calculated using a simulation method such as Monte Carlo or a probability of detection method, and second shape features corresponding to these sample interference peaks can be obtained and recorded in a sample data set.

[0109] In one possible implementation, a large number of sample object peaks under different conditions (e.g., different object reflectivities, different distances, different rain and fog scattering intensities, etc.) can be experimentally measured to obtain first shape features corresponding to these sample object peaks, and the first shape features corresponding to these sample object peaks can be recorded in a sample dataset. Furthermore, a large number of sample interference peaks under different conditions (e.g., different object reflectivities, different distances, different rain and fog scattering intensities, etc.) can be experimentally measured to obtain second shape features corresponding to these sample interference peaks, and the second shape features corresponding to these sample interference peaks can be recorded in a sample dataset.

[0110] Of course, the above is just an example of obtaining a sample data set, and there is no restriction on the source of the sample data set.

[0111] Step S12: For each parameter group in the support group, the configured typical value of the left peak width coefficient (such as an arbitrary value configured based on experience) is used as the left peak width coefficient in the parameter group, and the configured typical value of the right peak width coefficient (such as an arbitrary value configured based on experience) is used as the right peak width coefficient in the parameter group.

[0112] For example, for the first parameter group in the support group, that is, when the value of i is 1, the typical value of the left peak width coefficient 0 can be used as the left peak width coefficient in the parameter group, such as α1 is 0, and the typical value of the right peak width coefficient 1 can be used as the right peak width coefficient in the parameter group, such as β1 is 1. For the second parameter group in the support group, that is, when the value of i is 2, the typical value of the left peak width coefficient 1 can be used as the left peak width coefficient in the parameter group, such as α2 is 1, and the typical value of the right peak width coefficient 1 can be used as the right peak width coefficient in the parameter group, such as β2 is 1. For the third parameter group in the support group, that is, when the value of i is 3, the typical value of the left peak width coefficient 1 can be used as the left peak width coefficient in the parameter group, such as α3 is 1, and the typical value of the right peak width coefficient 0 can be used as the right peak width coefficient in the parameter group, such as β3 is 0. For the fourth parameter group in the support group, that is, when the value of i is 4, the typical value of the left peak width coefficient 0.3 can be used as the left peak width coefficient in the parameter group, such as α4 is 0.3, and the typical value of the right peak width coefficient 0.7 can be used as the right peak width coefficient in the parameter group, such as β4 is 0.7, and so on.

[0113] Of course, the above are just a few examples of the left peak width coefficient and the right peak width coefficient. There is no limitation on the left peak width coefficient and the right peak width coefficient, as long as the left peak width coefficient and the right peak width coefficient are not both zero. For example, the left peak width coefficient can be a value between 0 and 1, and the left peak width coefficient can be either 0 or 1. The right peak width coefficient can be a value between 0 and 1, and the right peak width coefficient can be either 0 or 1.

[0114] Furthermore, the ratio (ratio of the left peak width coefficient to the right peak width coefficient) may be different in different parameter groups.

[0115] Step S13: Based on the sample data set, for each calibration peak height, select a first shape feature corresponding to the calibration peak height and a second shape feature corresponding to the calibration peak height from the sample data set.

[0116] For example, multiple calibration peak heights can be obtained, which means that the peak height characteristic values corresponding to these calibration peak heights need to be recorded in the mapping table. There is no restriction on these calibration peak heights. The calibration peak heights can be pre-configured based on experience, or each peak height appearing in the sample data set can be used as the calibration peak height.

[0117] For each calibration peak height, since the first shape feature corresponding to the sample object peak includes peak height, left peak width, and right peak width, if the peak height in the first shape feature is the same as the calibration peak height, then the first shape feature is the first shape feature corresponding to the calibration peak height, thereby obtaining multiple first shape features corresponding to the calibration peak height. Since the second shape feature corresponding to the sample interference peak includes peak height, left peak width, and right peak width, if the peak height in the second shape feature is the same as the calibration peak height, then the second shape feature is the second shape feature corresponding to the calibration peak height, thereby obtaining multiple second shape features corresponding to the calibration peak height.

[0118] In summary, for each calibrated peak height, a plurality of first shape features corresponding to the calibrated peak height and a plurality of second shape features corresponding to the calibrated peak height may be selected from the sample data set.

[0119] Step S14: For each parameter group in the support group, based on the left peak width coefficient in the parameter group, the right peak width coefficient in the parameter group, the first shape feature corresponding to the calibrated peak height (i.e., each calibrated peak height), and the second shape feature corresponding to the calibrated peak height, determine the peak height characteristic value corresponding to the calibrated peak height, and record the correspondence between the calibrated peak height and the peak height characteristic value in the mapping table of the parameter group.

[0120] For example, for the first parameter group in the support group, based on the left peak width coefficient α1, the right peak width coefficient β1, the first shape feature and the second shape feature corresponding to the calibrated peak height p1, the peak height characteristic value corresponding to the calibrated peak height p1 is determined, and the corresponding relationship between the calibrated peak height p1 and the peak height characteristic value is recorded in the mapping table of the parameter group. Based on the left peak width coefficient α1, the right peak width coefficient β1, the first shape feature and the second shape feature corresponding to the calibrated peak height p2, the peak height characteristic value corresponding to the calibrated peak height p2 is determined, and the corresponding relationship between the calibrated peak height p2 and the peak height characteristic value is recorded in the mapping table of the parameter group, and so on.

[0121] For the second parameter group within the support group, based on the left peak width coefficient α2, the right peak width coefficient β2, the first shape feature and the second shape feature corresponding to the calibrated peak height p1, the peak height characteristic value corresponding to the calibrated peak height p1 is determined, and the corresponding relationship between the calibrated peak height p1 and the peak height characteristic value is recorded in the mapping table of the parameter group. Based on the left peak width coefficient α2, the right peak width coefficient β2, the first shape feature and the second shape feature corresponding to the calibrated peak height p2, the peak height characteristic value corresponding to the calibrated peak height p2 is determined, and the corresponding relationship between the calibrated peak height p2 and the peak height characteristic value is recorded in the mapping table of the parameter group, and so on.

[0122] Obviously, for each parameter group in the support group, the corresponding relationship between the calibrated peak height (ie, each calibrated peak height) and the peak height characteristic value can be recorded in the mapping table of the parameter group.

[0123] Exemplarily, when determining the peak height characteristic value, it is necessary to satisfy that the target characteristic value corresponding to each parameter group based on the sample object peak is less than a threshold value (such as 0), and the target characteristic value corresponding to each parameter group based on the sample interference peak is not less than a threshold value (such as 0). For example, for each calibration peak height (taking the calibration peak height p1 as an example), the left peak width coefficient α1 of the first parameter group, the left peak width in the first shape feature, the right peak width coefficient β1 of the first parameter group, and the right peak width in the first shape feature are substituted into expression (1), and it is necessary to satisfy f is less than 0. The left peak width coefficient α2 of the second parameter group, the left peak width in the first shape feature, the right peak width coefficient β2 of the second parameter group, and the right peak width in the first shape feature are substituted into expression (1), and it is necessary to satisfy f is less than 0, and so on. And, the left peak width coefficient α1 of the first parameter group, the left peak width in the second shape feature, the right peak width coefficient β1 of the first parameter group, and the right peak width in the second shape feature are substituted into expression (1), and it is necessary to satisfy f is not less than 0. Substitute the left peak width coefficient α2 of the second parameter group, the left peak width in the second shape feature, the right peak width coefficient β2 of the second parameter group, and the right peak width in the second shape feature into expression (1), and it is necessary to satisfy f is not less than 0, and so on. When the above conditions are met, the peak height characteristic value corresponding to the calibration peak height p1 is determined. That is, after substituting the peak height characteristic value into expression (1), the above conditions are met and the peak height characteristic value corresponding to the calibration peak height p1 in this case is solved.

[0124] Similarly, the peak height characteristic value corresponding to the calibration peak height p2 can be determined, ..., and so on.

[0125] For example, for each parameter group in the support group (taking parameter group i as an example), after obtaining the left peak width coefficient α of parameter group i i and right peak width coefficient β i Afterwards, by adjusting the peak height characteristic value γ i The value of (I) is such that the object peak region determined by the support group contains as many characteristic points corresponding to the object peak as possible (i.e., the first shape feature) and as few characteristic points corresponding to the interference peak as possible (i.e., the second shape feature). Based on the above principle, for each calibrated peak height, the peak height characteristic value corresponding to the calibrated peak height is determined in the following manner.

[0126] If the feature space region of the first shape feature corresponding to the sample object peak does not overlap with the feature space region of the second shape feature corresponding to the sample interference peak (for example, the feature space regions occupied by the two types of feature points do not overlap), the following expression (2) can be used to determine the peak height eigenvalue corresponding to the calibration peak height:

[0127]

[0128] In expression (2), αi represents the left peak width coefficient in parameter group i, β i represents the right peak width coefficient in parameter group i, can represent the left peak width, right peak width and peak height corresponding to the j-th sample object peak, respectively. They can represent the left peak width, right peak width and peak height corresponding to the j-th sample interference peak respectively; I′ represents the calibration peak height, γ i (I′) represents the peak height characteristic value corresponding to the calibrated peak height.

[0129] For example, for the calibration peak height p1, I' represents p1, then find all peaks with heights equal to the calibration peak height p1 from the sample data set (i.e. ), the first shape feature includes the left peak width corresponding to the sample object peak and right peak width The second shape feature includes the left peak width corresponding to the sample interference peak and right peak width Then the left peak width coefficient α i and right peak width coefficient β i Substituting into expression (2), we can get the peak height characteristic value corresponding to the calibrated peak height p1. In expression (2), It means taking the minimum value among the corresponding values of multiple sample object peaks. Indicates taking the maximum value among the corresponding values of multiple sample interference peaks.

[0130] In summary, the peak height characteristic value corresponding to each calibrated peak height can be obtained, and the corresponding relationship between the calibrated peak height and the peak height characteristic value is recorded in the mapping table corresponding to the parameter group i.

[0131] If the feature space region of the first shape feature corresponding to the sample object peak coincides with the feature space region of the second shape feature corresponding to the sample interference peak (e.g., the feature space regions occupied by the two types of feature points coincide, for example, if the laser pulse width is wide or the pile-up effect of the laser radar is severe, the feature space regions occupied by the two types of feature points coincide), then α i , β i and γ i The selection of (I) needs to include the minimum area of as many object peaks as possible. The peak height characteristic value corresponding to the calibration peak height can be determined using the following expression (3):

[0132]

[0133] In expression (3), α i represents the left peak width coefficient in parameter group i, β i represents the right peak width coefficient in parameter group i, It can represent the left peak width, right peak width and peak height corresponding to the j-th sample object peak, I′ represents the calibration peak height, γ i (I′) represents the peak height characteristic value corresponding to the calibrated peak height.

[0134] For example, for the calibration peak height p1, I' represents p1, then find all peaks with heights equal to the calibration peak height p1 from the sample data set (i.e. ), the first shape feature includes the left peak width corresponding to the sample object peak and right peak width Then the left peak width coefficient α i and right peak width coefficient β i Substituting into expression (3), we can get the peak height characteristic value corresponding to the calibrated peak height p1. In expression (3), Indicates taking the minimum value among the corresponding values of multiple sample object peaks.

[0135] In summary, the peak height characteristic value corresponding to each calibrated peak height can be obtained, and the corresponding relationship between the calibrated peak height and the peak height characteristic value is recorded in the mapping table corresponding to the parameter group i.

[0136] To summarize, for a given lidar, when the detector type and the emitted laser pulse waveform of the lidar are determined, preliminary calibration can be performed to determine the relevant parameters in each parameter group (such as the left peak width coefficient, the right peak width coefficient and the mapping table, and the mapping table includes the correspondence between the calibration peak height and the peak height eigenvalue), that is, determined by a training set containing a large number of feature points of known object peaks and interference peaks.

[0137] As can be seen from the above technical solutions, in the embodiment of the present application, by extracting the shape features (such as peak height, left peak width and right peak width) corresponding to the peak points of the histogram, the peak points are determined to be object peaks or interference peaks based on the shape features, and the peak points can be classified, thereby effectively avoiding the false alarms caused by rain and fog interference peaks when the laser radar is working in rain and fog scenes, so that the laser radar has the ability to classify the detected peaks, reduce the false alarms of the laser radar in rain and fog environments, filter out the interference of rain and fog and other environments, and ultimately achieve the purpose of improving the ranging accuracy. The ranging function can be realized normally, so that accurate ranging results can be obtained. A laser radar anti-rain and fog method based on peak shape can be implemented, and the three features of left peak width, right peak width and peak height are used to distinguish interference peaks from object peaks, and the result of whether the echo signal peak is an interference peak or an object peak is obtained. By extracting the three features of left peak width, right peak width and peak height on the histogram peak, the peaks are classified using the support set in the feature space, thereby effectively avoiding the false alarms caused by rain and fog interference peaks when the laser radar is working in rain and fog environments.

[0138] Based on the same application concept as the above method, a laser ranging device is proposed in the embodiment of the present application and applied to laser radar, see Figure 7 FIG. 1 is a schematic structural diagram of the device, which includes:

[0139] An acquisition module 71 is configured to acquire a histogram based on the detection data of the laser radar, wherein the abscissa of the histogram is the flight time of the laser pulse, and the ordinate of the histogram is the frequency corresponding to the flight time; and to acquire shape features corresponding to the peaks of the histogram, wherein the shape features include peak height, left peak width, and right peak width; wherein the peak height is the difference between the height of the highest point of the peak and the height of the background of the histogram, the left peak width is the width from the position of the highest point of the peak to the position where the height on the left side of the peak drops by a first height, and the right peak width is the width from the position of the highest point of the peak to the position where the height on the right side of the peak drops by a second height;

[0140] The determination module 72 is configured to determine whether the peak point is an object peak or an interference peak based on the shape feature; if the peak point is an object peak, determine the object distance based on the flight time corresponding to the peak point.

[0141] Exemplarily, when the determination module 72 determines that the peak point is an object peak or an interference peak based on the shape feature, it is specifically used to: determine a first eigenvalue based on the left peak width, determine a second eigenvalue based on the right peak width, and determine a third eigenvalue based on the peak height; the first eigenvalue and the second eigenvalue are both greater than or equal to 0, and the third eigenvalue is less than 0; if the left peak width is larger, the first eigenvalue is larger, if the right peak width is larger, the second eigenvalue is larger, and if the peak height is larger, the absolute value of the third eigenvalue is smaller; determine a target eigenvalue based on the first eigenvalue, the second eigenvalue and the third eigenvalue, and determine that the peak point is an object peak or an interference peak based on the target eigenvalue.

[0142] Exemplarily, the calibrated support group includes N parameter groups, where N is a positive integer, and each parameter group includes a left peak width coefficient, a right peak width coefficient and a mapping table, wherein the mapping table includes a correspondence between a calibrated peak height and a peak height eigenvalue; the determination module 72 is specifically used to determine whether the peak point is an object peak or an interference peak based on the shape feature: for each parameter group, determine a first eigenvalue based on the product value of the left peak width and the left peak width coefficient, determine a second eigenvalue based on the product value of the right peak width and the right peak width coefficient, and obtain a third eigenvalue corresponding to the peak height by querying the mapping table through the peak height; determine the target eigenvalue corresponding to the parameter group based on the first eigenvalue, the second eigenvalue and the third eigenvalue; if the target eigenvalue corresponding to each parameter group is less than a threshold value, determine the peak point as an object peak; if the target eigenvalue corresponding to any parameter group is not less than a threshold value, determine the peak point as an interference peak.

[0143] Exemplarily, for each parameter group in the support group, the determination module 72 is specifically used to calibrate the parameter group by: using the configured typical value of the left peak width coefficient as the left peak width coefficient in the parameter group, and using the configured typical value of the right peak width coefficient as the right peak width coefficient in the parameter group; obtaining a sample data set, the sample data set including a first shape feature corresponding to the sample object peak and a second shape feature corresponding to the sample interference peak; for each calibration peak height, selecting the first shape feature and the second shape feature corresponding to the calibration peak height from the sample data set; determining the peak height characteristic value corresponding to the calibration peak height based on the left peak width coefficient, the right peak width coefficient, the first shape feature and the second shape feature corresponding to the calibration peak height, and recording the correspondence between the calibration peak height and the peak height characteristic value in the mapping table of the parameter group.

[0144] Exemplarily, the determination module 72 determines the peak height characteristic value corresponding to the calibration peak height based on the left peak width coefficient, the right peak width coefficient, the first shape feature and the second shape feature corresponding to the calibration peak height: if the feature space region of the first shape feature corresponding to the sample object peak does not overlap with the feature space region of the second shape feature corresponding to the sample interference peak, then determine the peak height characteristic value using the following expression:

[0145] Alternatively, if the feature space region of the first shape feature corresponding to the sample object peak coincides with the feature space region of the second shape feature corresponding to the sample interference peak, the peak height feature value is determined using the following expression: Among them, α i represents the left peak width coefficient, β i represents the right peak width coefficient, They represent the left peak width, right peak width and peak height corresponding to the j-th sample object peak, respectively. Respectively represent the left peak width, right peak width and peak height corresponding to the j-th sample interference peak; I′ represents the calibration peak height, γ i (I′) represents the peak height characteristic value corresponding to the calibrated peak height.

[0146] Exemplarily, the acquisition module 71 is also used to obtain M statistical results corresponding to the peak point based on M consecutive histograms, where M is a positive integer, and for each statistical result, the statistical result indicates that the peak point is an object peak or an interference peak; the determination module 72 is also used to determine the number of occurrences of the object peak based on the M statistical results, and if the ratio of the number of occurrences of the object peak to M is greater than or equal to a first threshold, the peak point is determined to be an object peak; if the ratio of the number of occurrences of the object peak to M is less than the first threshold, the peak point is determined to be an interference peak; or, based on the M statistical results, determine the number of occurrences of the interference peak, and if the ratio of the number of occurrences of the interference peak to M is less than or equal to a second threshold, determine the peak point to be an object peak; if the ratio of the number of occurrences of the interference peak to M is greater than the second threshold, determine the peak point to be an interference peak.

[0147] Exemplarily, the laser ranging device is used to determine the distance of an object in a target scene, and the target scene includes a rain and fog scene; wherein the interference peak is an interference peak generated by rain and / or fog.

[0148] Based on the same application concept as the above method, an electronic device (such as an electronic device using a laser radar) is proposed in the embodiment of the present application, see Figure 8 As shown, it includes: a processor 81 and a machine-readable storage medium 82, the machine-readable storage medium 82 stores machine-executable instructions that can be executed by the processor 81; the processor 81 is used to execute the machine-executable instructions to implement the laser ranging method disclosed in the above example.

[0149] Based on the same application concept as the above method, an embodiment of the present application also provides a machine-readable storage medium, on which a number of computer instructions are stored. When the computer instructions are executed by a processor, the laser ranging method disclosed in the above example of the present application can be implemented.

[0150] The machine-readable storage medium may be any electronic, magnetic, optical, or other physical storage device that may contain or store information, such as executable instructions, data, and the like. For example, the machine-readable storage medium may be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, a storage drive (such as a hard disk drive), a solid-state drive, any type of storage disk (such as a CD, DVD, etc.), or similar storage media, or a combination thereof.

[0151] The systems, devices, modules, or units described in the above embodiments may be implemented by a computer or entity, or by a product having certain functions. A typical implementation device is a computer, which may be in the form of a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email transceiver, game console, tablet computer, wearable device, or any combination of these devices.

[0152] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0153] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the embodiments of the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0154] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0155] Furthermore, these computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0156] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0157] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A laser ranging method, characterized in that: Applied to laser radar, the method includes: A histogram is obtained based on the detection data of the laser radar, wherein the abscissa of the histogram is the flight time of the laser pulse, and the ordinate of the histogram is the frequency corresponding to the flight time; Obtaining shape features corresponding to the peak points of the histogram, the shape features including peak height, left peak width, and right peak width; wherein the peak height is the difference between the height of the highest point of the peak and the background height of the histogram, the left peak width is the width from the position of the highest point of the peak to the position where the height on the left side of the peak drops by a first height, and the right peak width is the width from the position of the highest point of the peak to the position where the height on the right side of the peak drops by a second height; determining, based on the shape feature, that the peak point is an object peak or an interference peak; If the peak point is an object peak, the object distance is determined based on the flight time corresponding to the peak point.

2. The method according to claim 1, characterized in that The determining, based on the shape feature, that the peak point is an object peak or an interference peak includes: Determine a first eigenvalue based on the left peak width, determine a second eigenvalue based on the right peak width, and determine a third eigenvalue based on the peak height; wherein both the first eigenvalue and the second eigenvalue are greater than or equal to 0, and the third eigenvalue is less than 0; the larger the left peak width, the larger the first eigenvalue, the larger the right peak width, the larger the second eigenvalue, and the larger the peak height, the smaller the absolute value of the third eigenvalue; A target eigenvalue is determined based on the first eigenvalue, the second eigenvalue, and the third eigenvalue, and the peak point is determined to be an object peak or an interference peak based on the target eigenvalue.

3. The method according to claim 1 or 2, characterized in that The calibrated support group includes N parameter groups, where N is a positive integer, and each parameter group includes a left peak width coefficient, a right peak width coefficient, and a mapping table, wherein the mapping table includes a correspondence between a calibrated peak height and a peak height characteristic value; The determining, based on the shape feature, that the peak point is an object peak or an interference peak includes: For each parameter group, a first eigenvalue is determined based on the product of the left peak width and the left peak width coefficient, a second eigenvalue is determined based on the product of the right peak width and the right peak width coefficient, and a third eigenvalue corresponding to the peak height is obtained by querying the mapping table through the peak height; a target eigenvalue corresponding to the parameter group is determined based on the first eigenvalue, the second eigenvalue, and the third eigenvalue; If the target characteristic values corresponding to each parameter group are all less than the threshold, the peak point is determined to be an object peak; If the target characteristic value corresponding to any parameter group is not less than the threshold, the peak point is determined to be an interference peak.

4. The method according to claim 3, characterized in that For each parameter group in the support group, the calibration process of the parameter group includes: Using the configured left peak width coefficient typical value as the left peak width coefficient in the parameter group, and using the configured right peak width coefficient typical value as the right peak width coefficient in the parameter group; Acquire a sample data set, the sample data set including a first shape feature corresponding to a sample object peak and a second shape feature corresponding to a sample interference peak; for each calibrated peak height, select the first shape feature and the second shape feature corresponding to the calibrated peak height from the sample data set; Based on the left peak width coefficient, the right peak width coefficient, the first shape feature and the second shape feature corresponding to the calibrated peak height, the peak height characteristic value corresponding to the calibrated peak height is determined, and the correspondence between the calibrated peak height and the peak height characteristic value is recorded in the mapping table of the parameter group.

5. The method according to claim 4, characterized in that Determining the peak height characteristic value corresponding to the calibrated peak height based on the left peak width coefficient, the right peak width coefficient, the first shape characteristic and the second shape characteristic corresponding to the calibrated peak height includes: If the feature space region of the first shape feature corresponding to the sample object peak does not overlap with the feature space region of the second shape feature corresponding to the sample interference peak, the following expression is used to determine the peak height eigenvalue: Alternatively, if the feature space region of the first shape feature corresponding to the sample object peak coincides with the feature space region of the second shape feature corresponding to the sample interference peak, the peak height feature value is determined using the following expression: Among them, α i represents the left peak width coefficient, β i represents the right peak width coefficient, They represent the left peak width, right peak width and peak height corresponding to the j-th sample object peak, respectively. They represent the left peak width, right peak width and peak height corresponding to the j-th sample interference peak respectively; I′ represents the calibration peak height, γ i (I′) represents the peak height characteristic value corresponding to the calibrated peak height.

6. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: Obtaining M statistical results corresponding to the peak point based on M consecutive histograms, where M is a positive integer, and for each statistical result, the statistical result indicates whether the peak point is an object peak or an interference peak; Determine the number of occurrences of the object peak based on the M statistical results; if the ratio of the number of occurrences of the object peak to M is greater than or equal to a first threshold, determine the peak point to be an object peak; if the ratio of the number of occurrences of the object peak to M is less than the first threshold, determine the peak point to be an interference peak; or The number of occurrences of the interference peak is determined based on the M statistical results. If the ratio of the number of occurrences of the interference peak to M is less than or equal to the second threshold, the peak point is determined to be an object peak. If the ratio of the number of occurrences of the interference peak to M is greater than the second threshold, the peak point is determined to be an interference peak.

7. The method according to any one of claims 1 to 5, characterized in that: The laser ranging method is used to determine the distance of an object in a target scene, wherein the target scene includes a rain and fog scene; The interference peak is caused by rain and / or fog.

8. A laser ranging device, characterized in that: Applied to laser radar, the device includes: an acquisition module, configured to acquire a histogram based on the detection data of the laser radar, wherein the abscissa of the histogram is the flight time of the laser pulse, and the ordinate of the histogram is the frequency corresponding to the flight time; and to acquire shape features corresponding to the peak points of the histogram, wherein the shape features include peak height, left peak width, and right peak width; wherein the peak height is the difference between the height of the highest point of the peak and the height of the background of the histogram, the left peak width is the width from the position of the highest point of the peak to the position where the height on the left side of the peak drops by a first height, and the right peak width is the width from the position of the highest point of the peak to the position where the height on the right side of the peak drops by a second height; A determination module is configured to determine whether the peak point is an object peak or an interference peak based on the shape feature; if the peak point is an object peak, determine the object distance based on the flight time corresponding to the peak point.

9. The device according to claim 8, It is characterized by: in, When the determination module determines that the peak point is an object peak or an interference peak based on the shape feature, it is specifically used to: determine a first eigenvalue based on the left peak width, determine a second eigenvalue based on the right peak width, and determine a third eigenvalue based on the peak height; wherein, the first eigenvalue and the second eigenvalue are both greater than or equal to 0, and the third eigenvalue is less than 0; if the left peak width is larger, the first eigenvalue is larger, if the right peak width is larger, the second eigenvalue is larger, and if the peak height is larger, the absolute value of the third eigenvalue is smaller; determine a target eigenvalue based on the first eigenvalue, the second eigenvalue, and the third eigenvalue, and determine that the peak point is an object peak or an interference peak based on the target eigenvalue; Among them, the calibrated support group includes N parameter groups, N is a positive integer, each parameter group includes a left peak width coefficient, a right peak width coefficient and a mapping table, the mapping table includes the correspondence between the calibrated peak height and the peak height eigenvalue; the determination module is specifically used to determine whether the peak point is an object peak or an interference peak based on the shape feature: for each parameter group, determine the first eigenvalue based on the product value of the left peak width and the left peak width coefficient, determine the second eigenvalue based on the product value of the right peak width and the right peak width coefficient, and obtain the third eigenvalue corresponding to the peak height by querying the mapping table through the peak height; determine the target eigenvalue corresponding to the parameter group based on the first eigenvalue, the second eigenvalue and the third eigenvalue; if the target eigenvalue corresponding to each parameter group is less than a threshold value, determine the peak point as an object peak; if the target eigenvalue corresponding to any parameter group is not less than a threshold value, determine the peak point as an interference peak; Wherein, for each parameter group in the support group, the determination module is specifically used to calibrate the parameter group: use the configured left peak width coefficient typical value as the left peak width coefficient in the parameter group, and use the configured right peak width coefficient typical value as the right peak width coefficient in the parameter group; obtain a sample data set, the sample data set including a first shape feature corresponding to a sample object peak and a second shape feature corresponding to a sample interference peak; for each calibration peak height, select the first shape feature and the second shape feature corresponding to the calibration peak height from the sample data set; based on the left peak width coefficient, the right peak width coefficient, the first shape feature and the second shape feature corresponding to the calibration peak height, determine the peak height characteristic value corresponding to the calibration peak height, and record the corresponding relationship between the calibration peak height and the peak height characteristic value in the mapping table of the parameter group; The determination module determines the peak height characteristic value corresponding to the calibration peak height based on the left peak width coefficient, the right peak width coefficient, the first shape feature and the second shape feature corresponding to the calibration peak height. If the feature space region of the first shape feature corresponding to the sample object peak does not overlap with the feature space region of the second shape feature corresponding to the sample interference peak, the peak height characteristic value is determined using the following expression: Alternatively, if the feature space region of the first shape feature corresponding to the sample object peak coincides with the feature space region of the second shape feature corresponding to the sample interference peak, the peak height feature value is determined using the following expression: Among them, α i represents the left peak width coefficient, β i represents the right peak width coefficient, They represent the left peak width, right peak width and peak height corresponding to the j-th sample object peak, respectively. They represent the left peak width, right peak width and peak height corresponding to the j-th sample interference peak respectively; I′ represents the calibration peak height, γ i (I′) represents the peak height characteristic value corresponding to the calibration peak height; Wherein, the acquisition module is further used to obtain M statistical results corresponding to the peak point based on M consecutive histograms, where M is a positive integer, and for each statistical result, the statistical result indicates that the peak point is an object peak or an interference peak; the determination module is further used to determine the number of occurrences of the object peak based on the M statistical results, and if the ratio of the number of occurrences of the object peak to M is greater than or equal to a first threshold, the peak point is determined to be an object peak; if the ratio of the number of occurrences of the object peak to M is less than the first threshold, the peak point is determined to be an interference peak; or, based on the M statistical results, determine the number of occurrences of the interference peak, and if the ratio of the number of occurrences of the interference peak to M is less than or equal to a second threshold, the peak point is determined to be an object peak; if the ratio of the number of occurrences of the interference peak to M is greater than the second threshold, the peak point is determined to be an interference peak; The laser ranging device is used to determine the distance of an object in a target scene, and the target scene includes a rain and fog scene; wherein the interference peak is an interference peak generated by rain and / or fog.

10. An electronic device, characterized in that: include: a processor and a machine-readable storage medium storing machine-executable instructions capable of being executed by the processor; The processor is configured to execute machine-executable instructions to implement the method according to any one of claims 1 to 7.

Images