A laser pulse radar device, a method for improving point cloud density, and a device

Abstract

The present invention provides a method and device for improving the point cloud density of a laser pulse radar. The method for improving the point cloud density of the laser pulse radar includes: obtaining the return signals received by a low-frequency laser pulse radar and a high-frequency laser pulse radar, and determining the scanning interval times of the low-frequency laser pulse radar and the high-frequency laser pulse radar. Using the return signal received by the low-frequency laser pulse radar to correct and fit the return signal received by the high-frequency laser pulse radar to form point cloud data. By performing interval scanning with the low-frequency laser pulse radar and the high-frequency laser pulse radar, and through the correction and fitting process, correcting the positions of the point cloud information obtained by the high-frequency laser pulse radar based on the point cloud information obtained by the low-frequency laser pulse radar, the present invention effectively improves the point cloud density.

Description

Technical Field

[0001] The present invention pertains to the field of lidar for autonomous driving, involving technologies related to laser scanning and data processing. The main objective is to enhance the point cloud density. Specifically, it relates to a method and device for improving the point cloud density of a lidar. Background Art

[0002] With the development of the intelligent era, lidar, as an important 3D sensor, plays a crucial role especially in the field of autonomous driving. Currently, many vehicles in the automotive industry are equipped with lidar to assist in vehicle driving. Existing lidars for assisting vehicle driving mostly adopt hybrid solid-state MEMS lidars. The hybrid solid-state MEMS lidar itself measures the distance by the flight time that the laser beam experiences from being emitted through the vibration deflection of the micromirror to the return of the laser beam. After multiple vibrations and deflections of the galvanometer mirror, a complete three-dimensional spatial stereoscopic data with a certain viewing angle is collected. Generally, for the laser emitted through the vibration of the same micromirror, only one frequency of laser is selected for emission. High-frequency lasers have the advantages of short flight time and more object information measured, but the measured distance is relatively shorter than that of low-frequency lasers. If used to detect distant objects, distant objects will be misjudged as being closer, resulting in incorrect detection of the distance. Low-frequency lasers are the opposite. When scanning distant objects, only low-frequency lasers can be selected for scanning, and the object information returned by the low-frequency laser scanning is relatively less, and some may be filtered out as noise points, making it difficult to judge the reality of the object. The lack of object point cloud data has a negative impact on the target judgment of vehicle assisted driving. Summary of the Invention

[0003] The present invention provides a method for improving the point cloud density of a lidar to increase the object information that can be obtained by the lidar.

[0004] The present invention provides a method for improving the point cloud density of a lidar, and the method for improving the point cloud density of the lidar includes:

[0005] Obtain the return signals received by the low-frequency lidar and the high-frequency lidar, and determine the scanning interval times of the low-frequency lidar and the high-frequency lidar. Use the return signal received by the low-frequency lidar to correct and fit the return signal received by the high-frequency lidar to form point cloud data.

[0006] Further, the method for improving the point cloud density of the lidar is specifically as follows:

[0007] S101: The low-frequency lidar and the high-frequency lidar scan at intervals of a unit time, respectively based on the same micromirror.

[0008] S102: Obtain the return signals received by the low-frequency laser pulse radar and the high-frequency laser pulse radar, with the unit time as the scanning interval time;

[0009] S103: Compare the data of the high-frequency laser line with the distance of the low-frequency laser line, and calculate the delay reception time of the high-frequency laser pulse radar;

[0010] S104: Eliminate the misalignment of the high-frequency laser pulse radar signals, perform fitting, and improve the point cloud density.

[0011] Furthermore, the S101 is specifically: Based on the same laser source and micro-vibrating mirror, the low-frequency laser pulse radar and the high-frequency laser pulse radar are respectively scanned in an alternating odd-even line manner.

[0012] Furthermore, the S103 is specifically: Compare the reception signal periods of the two frequencies of high-frequency laser and low-frequency laser, and calculate the delay reception time period of the high-frequency laser pulse radar.

[0013] Furthermore, the delay reception time period is the reception signal period of the low-frequency laser minus the reception signal period of the high-frequency laser.

[0014] Furthermore, the S104 includes:

[0015] S1041. Determine the object position according to the point cloud information of the low-frequency laser pulse radar;

[0016] S1042. Search for the abnormal delay point cloud data of the surrounding adjacent high-frequency scans, and determine the location where the abnormal delay point cloud data is located;

[0017] S1043. Determine the offset distance of the abnormal delay point cloud data according to the delay reception time period and the light flight speed;

[0018] S1044. Fit the abnormal delay point cloud data on the point cloud information of the low-frequency laser pulse radar to obtain the corrected point cloud information.

[0019] On the other hand, the present invention discloses a laser pulse radar device, and the laser pulse radar device applies the above-mentioned laser pulse radar point cloud density improvement method.

[0020] On the other hand, the present invention also discloses a computer-readable storage medium, and the computer-readable storage medium contains computer program instructions,

[0021] When the computer program instructions are called, they are used to execute the above-mentioned laser pulse radar point cloud density improvement method.

[0022] Compared with the prior art, the present invention performs interval scanning with a low-frequency laser pulse radar and a high-frequency laser pulse radar, and through a correction and fitting process, corrects the position of the point cloud information obtained by the high-frequency laser pulse radar based on the point cloud information obtained by the low-frequency laser pulse radar, effectively improving the point cloud density. Description of the Drawings

[0023] Figure 1 It is a flowchart of the method for improving the point cloud density of the laser pulse radar according to the embodiment of the present invention;

[0024] Figure 2 It is the point cloud information collected according to the embodiment of the present invention;

[0025] Figure 3 It is the corrected point cloud information according to the embodiment of the present invention. Detailed Embodiments

[0026] In order to enable those skilled in the art of the present technology to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence.

[0028] The present invention provides a method for improving the point cloud density of a laser pulse radar. The method for improving the point cloud density of the laser pulse radar includes:

[0029] Obtain the return signals received by the low-frequency laser pulse radar and the high-frequency laser pulse radar, and determine the scanning interval time of the low-frequency laser pulse radar and the high-frequency laser pulse radar. Use the return signal received by the low-frequency laser pulse radar to correct and fit the return signal received by the high-frequency laser pulse radar to form point cloud data.

[0030] Among them, the method can be to first emit a low-frequency laser pulse radar and then emit a high-frequency laser pulse radar, or to first emit a high-frequency laser pulse radar and then emit a low-frequency laser pulse radar. After obtaining the return signals received by the low-frequency laser pulse radar and the high-frequency laser pulse radar, point cloud information is formed. According to the point cloud information received by the low-frequency laser pulse radar, the point cloud information received by the high-frequency laser pulse radar is corrected, so that the point cloud information collected by the high-frequency laser pulse radar can be redistributed with reference to the point cloud information collected by the low-frequency laser pulse radar to form a high-density point cloud.

[0031] In the embodiment of the present invention, the low-frequency laser pulse radar and the high-frequency laser pulse radar perform interval scanning, and through the correction fitting process, the position of the point cloud information obtained by the high-frequency laser pulse radar is corrected based on the point cloud information obtained by the low-frequency laser pulse radar, effectively improving the point cloud density.

[0032] Optionally, as Figure 1 shown, the method for improving the point cloud density of the laser pulse radar is specifically as follows:

[0033] S101: The low-frequency laser pulse radar and the high-frequency laser pulse radar are separated by a unit time and scan respectively based on the same micro-vibrating mirror;

[0034] Among them, the devices where the low-frequency laser pulse radar and the high-frequency laser pulse radar are located are the same laser source and MEMS, and the two laser frequencies are emitted and scanned in an adjacent-row and alternating manner;

[0035] S102: Obtain the return signals received by the low-frequency laser pulse radar and the high-frequency laser pulse radar, and use the unit time as the scanning interval time;

[0036] S103: Compare the data of the high-frequency laser row with the distance of the low-frequency laser row, and calculate the delay reception time of the high-frequency laser pulse radar;

[0037] S104: Eliminate the misalignment of the high-frequency laser pulse radar signal, perform fitting, and improve the point cloud density.

[0038] Among them, the point cloud data of the two frequencies are searched and verified with each other, and the point cloud at the delay offset position in the high-frequency laser pulse radar signal is moved to the correct position.

[0039] Particularly, the S101 is specifically: Based on the same laser source and micro-vibrating mirror, the low-frequency laser pulse radar and the high-frequency laser pulse radar are scanned in an odd-even row alternating manner.

[0040] Among them, in the embodiments of the present invention, based on the same laser source, while the galvanometer vibrates, the laser is controlled to sequentially emit two different frequencies of high and low in an adjacent row and alternating manner. When the low frequency is emitted, the time interval is long, and the actually detectable distance is farther. The data obtained, regardless of the distance, are correct distance values. However, the point cloud density of the detected effective object is low; for the high-frequency laser, the time interval is short, and the detectable distance is short. There may be an echo from a distant object emitted last time, resulting in incorrect distance detection, moving the distant object to the near, but the point cloud density of the effective object is high.

[0041] In the embodiments of the present invention, by adopting the adjacent row scanning method, it can be ensured that there are high-frequency and low-frequency laser pulse radars for the same object being scanned, which is convenient for searching the true position of the effective object in the delay situation generated when the high-frequency laser is used to detect distant objects based on the low-frequency laser and subsequent algorithms.

[0042] Specifically, S103 is: comparing the received signal periods of the high-frequency laser and the low-frequency laser, and calculating the delay reception time period of the high-frequency laser pulse radar.

[0043] Specifically, the delay reception time period is the received signal period of the low-frequency laser minus the received signal period of the high-frequency laser.

[0044] Among them, when transmitting and receiving with two frequencies, the low frequency can normally receive the returned laser pulse radar signal, while the high-frequency laser has a limited time interval for detecting distant objects, so it cannot receive the returned laser pulse radar signal within the effective time, that is, there will be a certain delay in reception. That is, when receiving the echo of the current high-frequency laser emission, part of the echo of the previous high-frequency laser emission will be received at the same time. Since the system will recognize part of the echo of the previous high-frequency laser emission as the echo of the current high-frequency laser emission, it will cause the dislocation of the point cloud information in part of the echo of the previous high-frequency laser emission.

[0045] In the embodiments of the present invention, by obtaining the delay reception time period, the part of the echo of the previous high-frequency laser emission can be corrected. After determining which point cloud information belongs to the echo of the previous high-frequency laser emission, by determining the delay reception time period, the correct position of the point cloud information can be obtained through calculation to achieve the correction effect.

[0046] Specifically, as Figure 2 , 3 shown, S104 includes:

[0047] S1041. Determining the object position according to the point cloud information of the low-frequency laser pulse radar;

[0048] Among them, the returned signal of the low-frequency laser pulse radar is analyzed to obtain accurate point cloud information, and the object position is determined according to the point cloud information.

[0049] S1042. Search for abnormal delayed point cloud data of adjacent high-frequency scans around, and determine the location where the abnormal delayed point cloud data is located;

[0050] Among them, by comparing and analyzing the point cloud information of the high-frequency lidar and the point cloud information of the low-frequency lidar, if the high-frequency lidar has a large amount of point cloud information at a certain position while the low-frequency lidar does not have point cloud information at that position, it is determined that the point cloud information of the high-frequency lidar at that position is abnormal; as Figure 2 shown, the circled position is the offset point cloud position of the high-frequency lidar;

[0051] S1043. Determine the offset distance of the abnormal delayed point cloud data according to the delayed reception time period and the light flight speed;

[0052] Among them, the product of the delayed reception time period and the light flight speed is the offset distance;

[0053] S1044. Fit the abnormal delayed point cloud data to the point cloud information of the low-frequency lidar to obtain the corrected point cloud information.

[0054] Among them, move the abnormal delayed point cloud data along the emission direction of the lidar according to the offset distance, and fit it to the point cloud information of the low-frequency lidar to obtain the corrected point cloud information. As Figure 3 shown, move the offset position to the correct position through software algorithms, so that the low-frequency point cloud position and the high-frequency point cloud position are combined together, improving the point cloud density.

[0055] In the embodiment of the present invention, to eliminate the delay error generated by high-frequency laser detecting distant objects, first determine the specific position of the object through the low-frequency lidar. Since the high-frequency and low-frequency lasers scan alternately in odd and even rows, and the low-frequency lidar returns normally, that is, the point cloud position of the low-frequency lidar is accurate. After the same object is scanned by the high-frequency and low-frequency lidars, the formed point cloud information should be adjacent (that is, the low-frequency lidar has a small amount of point cloud at the position of the object, and the high-frequency lidar will necessarily have a large amount of point cloud information at the position of the object). Therefore, by using a correction software algorithm, the abnormal delayed point cloud data scanned by the high-frequency lidar can be searched through the point cloud data scanned by the low-frequency lidar, and its location can be determined. Finally, add the offset distance obtained by multiplying the delay time by the light flight speed to determine its correct position. In this way, the point cloud density can be improved while ensuring that the low-frequency detection can detect objects at a farther distance.

[0056] The multiple situation of improving the point cloud density is as follows:

[0057] Let the low frequency be h, the high frequency be s, and the multiple k = (h / 2 + s / 2) / h;

[0058] If h = 500 kHz and s = 1000 kHz, then k = 1.5, that is, the point cloud density is 1.5 times the original.

[0059] Among them, the high and low frequencies of the low-frequency laser pulse radar and the high-frequency laser pulse radar are relative, not absolute. The user can adjust the frequencies of the specific low-frequency laser pulse radar and high-frequency laser pulse radar according to the solution of the embodiment of the present invention based on specific detection requirements. In the embodiment of the present invention, increasing the frequency of the high-frequency laser pulse radar can increase the point cloud density, and decreasing the frequency of the low-frequency laser pulse radar can increase the detection at a long distance. Different models of radars can also be applicable to the current scanning method.

[0060] On the other hand, the present invention discloses a laser pulse radar device that applies the above method for increasing the point cloud density of the laser pulse radar.

[0061] Among them, the laser pulse radar device is a device with the detection functions of a low-frequency and a high-frequency laser pulse radar, and this device has the same laser source and micro-vibrating mirror.

[0062] On the other hand, the present invention also discloses a computer-readable storage medium that contains computer program instructions.

[0063] When the computer program instructions are called, they are used to execute the above method for increasing the point cloud density of the laser pulse radar.

[0064] Among them, the computer-readable storage medium can be optionally a USB flash drive, a cloud disk, a hard disk, and other electronic devices or cloud platforms with storage functions.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that after reading the specification of this application, they can still modify the specific implementation manners of the present invention or make equivalent replacements, but these modifications or changes are all within the protection scope of the pending claims of this invention application.

Claims

1. A method for improving the point cloud density of a lidar pulse radar, characterized in that, The method for improving the point cloud density of the lidar includes: Obtain the return signals received by the low-frequency lidar and the high-frequency lidar, and determine the scanning interval times of the low-frequency lidar and the high-frequency lidar. Use the return signal received by the low-frequency lidar to correct and fit the return signal received by the high-frequency lidar to form point cloud data; The method for improving the point cloud density of the lidar is specifically as follows: S101: The low-frequency lidar and the high-frequency lidar scan at intervals of a unit time, respectively, based on the same micro-mirror; S102: Obtain the return signals received by the low-frequency lidar and the high-frequency lidar, and use the unit time as the scanning interval time; S103: Compare the data of the high-frequency laser line with the distance of the low-frequency laser line, and calculate the delay reception time of the high-frequency lidar; S104: Eliminate the dislocation of the high-frequency lidar signal, perform fitting, and improve the point cloud density.

2. The method for improving the point cloud density of a lidar according to claim 1, characterized in that The specific content of S101 is: Based on the same laser source and micro-mirror, the low-frequency lidar and the high-frequency lidar scan in an alternating odd-even line manner.

3. The method for improving the point cloud density of a lidar according to claim 1, wherein The specific content of S103 is: Compare the reception signal periods of the high-frequency laser and the low-frequency laser, and calculate the delay reception time period of the high-frequency lidar.

4. The method for improving the point cloud density of a lidar according to claim 3, wherein The delay reception time period is the reception signal period of the low-frequency laser minus the reception signal period of the high-frequency laser.

5. A method for improving the point cloud density of a lidar pulse radar according to claim 1, characterized in that, The S104 includes: S1041. Determine the object position according to the point cloud information of the low-frequency lidar; S1042. Search for the abnormal delay point cloud data of the surrounding adjacent high-frequency scans, and determine the position where the abnormal delay point cloud data is located; S1043. Determine the offset distance of the abnormal delay point cloud data according to the delay reception time period and the light flight speed; S1044. Fit the abnormal delay point cloud data to the point cloud information of the low-frequency lidar to obtain the corrected point cloud information.

6. A laser pulse radar device, characterized in that, The lidar device applies the method for improving the point cloud density of the lidar according to any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains computer program instructions, which are used to execute the method for improving the point cloud density of the lidar according to any one of claims 1-5 when the computer program instructions are called.

Images