Dirt detection device and method, storage medium, laser radar and vehicle

Abstract

The invention relates to a dirt detection device and method, a storage medium, a laser radar and a vehicle. The device comprises a laser radar window sheet, a control unit, an echo signal receiving unit and an auxiliary light source, wherein the echo signal receiving unit and the auxiliary light source are respectively connected with the control unit; wherein the auxiliary light source and the echo signal receiving unit are arranged in the same cavity; the auxiliary light source is configured to emit a detection light signal to the laser radar window sheet; the echo signal receiving unit is configured to receive a reflected light signal of the detection light signal reflected by the laser radar window sheet, and then determine a target image of the laser radar window sheet according to the reflected light signal; and the control unit is configured to perform smudginess detection on the laser radar window sheet according to the target image and a preset reference image.

Description

Technical Field

[0001] The present disclosure relates to the field of lidar technology, and in particular, to a dirt detection device, method, storage medium, lidar, and vehicle. Background Art

[0002] Lidar is an important sensor for current autonomous vehicles. When an autonomous vehicle is driving on outdoor road conditions, it may be exposed to different dirt such as rain, snow, sand, sewage splashing, or insect residues, which may adhere to the window pane of the lidar sensor mounted on the vehicle. This will cause the lidar to be unable to accurately obtain the surrounding road conditions, resulting in inaccurate and untimely distance information for detecting the surrounding environment of the vehicle, thereby affecting driving safety. Summary of the Invention

[0003] To overcome the problems existing in the related art, the present disclosure provides a dirt detection device, method, storage medium, lidar, and vehicle.

[0004] According to a first aspect of an embodiment of the present disclosure, a dirt detection device is provided, including:

[0005] A lidar window pane, a control unit, an echo signal receiving unit and an auxiliary light source respectively connected to the control unit; wherein, the auxiliary light source and the echo signal receiving unit are arranged in the same cavity;

[0006] The auxiliary light source is configured to emit a detection optical signal to the lidar window pane;

[0007] The echo signal receiving unit is configured to determine a target image of the lidar window pane according to the reflected optical signal after receiving the reflected optical signal of the detection optical signal reflected by the lidar window pane;

[0008] The control unit is configured to perform dirt detection on the lidar window pane according to the target image and a preset reference image.

[0009] Optionally, the control unit is configured to subtract the gray value of each pixel point of the target image and the preset reference image to obtain a dirt image; perform a histogram statistics on the dirt image to obtain a dirt detection result, where the dirt detection result includes at least one target gray value corresponding to the pixel points of the dirt and the number of pixel points corresponding to each target gray value.

[0010] Optionally, the control unit is configured to, for each first pixel point of the target image, obtain a gray difference value between the gray value of the first pixel point of the target image and the gray value of a second pixel point on the preset reference image, where the second pixel point is the pixel point on the preset reference image corresponding to the first pixel point; use the first pixel points with the gray difference value greater than or equal to a preset difference threshold as third pixel points; replace the gray value of the third pixel points with the gray difference value, and set the gray values of other pixel points except the third pixel points to a preset gray value, so as to obtain the dirty image.

[0011] Optionally, the control unit is configured to, for each gray value corresponding to the dirty image, determine the occurrence probability of the gray value in the dirty image; perform a normalization process on the dirty image so as to normalize the gray value of each pixel point in the dirty image to a preset gray value range; determine at least one target gray value corresponding to the dirty pixel points in the preset gray value range according to the occurrence probability, and the number of pixel points corresponding to each target gray value.

[0012] Optionally, the echo signal receiving unit is configured to, after converting the reflected light signal into an electrical signal, perform analog-to-digital conversion according to the electrical signal, and then determine the target image.

[0013] Optionally, the control unit is further configured to obtain the current exposure rate of the echo signal receiving unit, and control the emission power of the detection light signal emitted by the auxiliary light source according to the exposure rate.

[0014] Optionally, the device further includes: a dirty processing execution unit connected to the control unit;

[0015] The control unit is further configured to send a dirty detection result to the dirty processing execution unit;

[0016] The dirty processing execution unit is configured to perform a preset dirty cleaning operation on the lidar window according to the dirty detection result.

[0017] Optionally, a laser emitting unit;

[0018] The control unit is further configured to send a notification message indicating that the dirty cleaning is completed to the laser emitting unit;

[0019] The laser emitting unit is configured to emit a laser signal outward after receiving the notification message.

[0020] According to a second aspect of the embodiments of the present disclosure, a dirt detection method is provided, which is applied to a dirt detection device. The device includes: a lidar window, a control unit, an echo signal receiving unit and an auxiliary light source respectively connected to the control unit; wherein, the auxiliary light source and the echo signal receiving unit are arranged in the same cavity; the method includes:

[0021] Transmit a detection optical signal to the lidar window through the auxiliary light source;

[0022] After the echo signal receiving unit receives the reflected optical signal of the detection optical signal reflected by the lidar window, determine the target image of the lidar window according to the reflected optical signal;

[0023] Perform dirt detection on the lidar window through the control unit according to the target image and a preset reference image.

[0024] Optionally, the performing dirt detection on the lidar window through the control unit according to the target image and a preset reference image includes:

[0025] Subtract the gray values of each pixel point of the target image and the preset reference image to obtain a dirt image;

[0026] Perform histogram statistics on the dirt image to obtain a dirt detection result, where the dirt detection result includes at least one target gray value corresponding to the pixel points of the dirt and the number of pixel points corresponding to each target gray value.

[0027] Optionally, the subtracting the gray values of each pixel point of the target image and the preset reference image to obtain a dirt image includes:

[0028] For each first pixel point of the target image, obtain the gray value difference between the gray value of the first pixel point of the target image and the gray value of the second pixel point on the preset reference image, where the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image;

[0029] Use the first pixel points whose gray value difference is greater than or equal to a preset difference threshold as third pixel points;

[0030] Replace the gray value of the third pixel point with the gray value difference, and set the gray values of other pixel points except the third pixel point to a preset gray value to obtain the dirt image.

[0031] Optionally, the performing histogram statistics on the dirt image to obtain a dirt detection result includes:

[0032] For each gray value corresponding to the dirty image, determine the occurrence probability of the gray value in the dirty image;

[0033] Perform normalization processing on the dirty image so as to normalize the gray value of each pixel point in the dirty image to a preset gray value interval;

[0034] Determine at least one target gray value corresponding to the dirty pixel points in the preset gray value interval according to the occurrence probability, and the number of pixel points corresponding to each target gray value.

[0035] Optionally, the determining the target image of the lidar window sheet according to the reflected light signal includes:

[0036] Convert the reflected light signal into an electrical signal;

[0037] After performing analog-to-digital conversion according to the electrical signal, determine the target image.

[0038] Optionally, the method further includes:

[0039] Obtain the current exposure rate of the echo signal receiving unit through the control unit, and control the emission power of the detection light signal emitted by the auxiliary light source according to the exposure rate.

[0040] Optionally, the device further includes: a dirty processing execution unit connected to the control unit; the method further includes:

[0041] Send the dirty detection result to the dirty processing execution unit through the control unit;

[0042] According to the dirty detection result, perform a preset dirty cleaning operation on the lidar window sheet through the dirty processing execution unit.

[0043] Optionally, a laser emitting unit; the method further includes:

[0044] Send a notification message indicating that the dirty cleaning is completed to the laser emitting unit through the control unit;

[0045] After receiving the notification message, the laser emitting unit emits a laser signal outward.

[0046] According to a third aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium, on which computer program instructions are stored, and when the program instructions are executed by a processor, the steps of the dirty detection method provided in the second aspect of the present disclosure are implemented.

[0047] According to a fourth aspect of the embodiments of the present disclosure, there is provided a lidar, including the dirty detection device described in the first aspect of the present disclosure.

[0048] According to a fifth aspect of the embodiments of the present disclosure, there is provided a vehicle, including the lidar described in the fourth aspect of the present disclosure.

[0049] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects: By providing an auxiliary light source in the receiving cavity of the lidar (i.e., the receiving cavity where the echo signal receiving unit is located), the contamination detection of the lidar window can be achieved, thus avoiding the problem that the contamination detection depends on the presence of targets in the external environment. Moreover, compared with the solution of adding a CMOS or CCD optoelectronic sensor inside the lidar to achieve contamination detection without targets, since most of the auxiliary light sources can select infrared light sources with lower costs, the present disclosure can reduce the detection cost by providing an auxiliary light source to achieve the contamination detection of the lidar window.

[0050] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present disclosure and, together with the specification, are used to explain the principles of the present disclosure.

[0052] Figure 1 is a structural block diagram of a contamination detection device shown according to an exemplary embodiment.

[0053] Figure 2 is a schematic diagram of the propagation of optical signals for lidar contamination detection shown according to an exemplary embodiment.

[0054] Figure 3 is a schematic diagram of a contamination histogram curve shown according to an exemplary embodiment.

[0055] Figure 4 is according to Figure 1 shown in the embodiment is a structural block diagram of a contamination detection device.

[0056] Figure 5 is a flowchart of a contamination detection method shown according to an exemplary embodiment.

[0057] Figure 6 is according to Figure 5 shown in the embodiment is a flowchart of a contamination detection method.

[0058] Figure 7 is according to Figure 5 shown in the embodiment is a flowchart of a contamination detection method.

[0059] Figure 8 is according toFigure 7 Flowchart of a dirt detection method shown in the illustrated embodiment.

[0060] Figure 9 Block diagram of a lidar shown according to an exemplary embodiment.

[0061] Figure 10 Block diagram of a vehicle shown according to an exemplary embodiment. Detailed implementation

[0062] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0063] It should be noted that all actions of obtaining signals, information, or data in this application are carried out on the premise of complying with the corresponding data protection regulations and policies of the country where the location is located and obtaining the authorization given by the owner of the corresponding device.

[0064] The present disclosure is mainly applied to the scenario of detecting dirt on the window pane of a lidar. After the window pane of the lidar is attached with dirt, the lidar sensor cannot accurately receive the echo energy emitted by the laser (for example, a part of the echo energy is lost), resulting in the distance calculation process of the control processing circuit not being able to accurately reflect the true distance between the lidar and the target object.

[0065] In the dirt detection methods provided in the related art, one is to use the own light source of the lidar and its own SiPMs (Silicon Photo Multipliers) infrared sensor to perform local dirt detection on the window pane on the premise that there is a target object to be measured in the surrounding environment (when there is a target object in the surrounding environment, the reflected light can enter the internal sensor of the lidar). However, this method requires the presence of a target object in the surrounding environment, and when there is no target object in the surrounding environment, dirt detection cannot be performed.

[0066] In another implementation, by adding a CMOS (Complementary Metal Oxide Semiconductor) or a CCD (Charge Coupled Devices) optoelectronic sensor inside the lidar, global dirt detection of the window pane without a target object is performed, but this requires adding an additional CMOS or CCD sensor, which will obviously increase the manufacturing cost of the lidar.

[0067] In addition, the related technology also provides a detection method based on scanning laser radar to receive the electrical signal of the laser field area and judge the threshold of the received electrical signal point by point. However, in the future, laser radar sensor chips will be integrated, and such a detection method will not only increase the cost of sensor chips sharply, but also need to improve the detection efficiency and reliability.

[0068] In order to solve the above problems, the present invention provides a dirt detection device, method, storage medium, laser radar and vehicle. The specific implementation of the present invention is described in detail below with reference to the accompanying drawings.

[0069] Figure 1 is a structural block diagram of a dirt detection device according to an exemplary embodiment. Figure 1 As shown, the device 100 includes:

[0070] The laser radar window sheet 101, the control unit 102, the echo signal receiving unit 103 and the auxiliary light source 104 connected to the control unit 102 respectively; wherein the auxiliary light source 104 and the echo signal receiving unit 103 are arranged in the same cavity. Figure 1 As shown, the auxiliary light source 104 and the echo signal receiving unit 103 are both arranged in the receiving cavity of the laser radar.

[0071] In one embodiment, the auxiliary light source 104 can use a low-cost infrared LED device, or a laser transmitter. It should be noted that the wavelength of the light signal emitted by the auxiliary light source needs to be the same as the wavelength of the laser emitted by the laser transmitter of the laser radar. In addition, when setting the auxiliary light source 104, the installation angle of the auxiliary light source 104 needs to be arranged as a target angle, wherein the target angle is the angle at which the light detection signal emitted by the auxiliary light source 104 can cover the entire optical receiving area window plate (usually refers to the area of ​​the laser radar window plate 101 facing the laser radar receiving cavity).

[0072] In addition, the echo signal receiving unit 103 may include a laser radar sensor located in a laser radar receiving cavity for receiving echo energy emitted by a laser transmitter (located in a laser radar transmitting cavity).

[0073] In this way, the auxiliary light source 104 can be configured to emit a detection optical signal towards the lidar window 101; the echo signal receiving unit 103 can be configured to receive the reflected optical signal of the detection optical signal reflected by the lidar window 101 and determine the target image of the lidar window 101 according to the reflected optical signal; the control unit 102 can be configured to perform dirt detection on the lidar window 101 according to the target image and a preset reference image. Wherein, the preset reference image can include a pre-set clean image when the lidar window 101 has no dirt.

[0074] Exemplarily, Figure 2 FIG. is a schematic diagram of the propagation of an optical signal for lidar dirt detection shown according to an exemplary embodiment, wherein, Figure 2 the solid arrows in represent the propagation route of the laser signal emitted by the lidar based on the VCSEL (Vertical-Cavity Surface-Emitting Laser) sensor as the laser emitter, Figure 2 the dashed arrows in represent the propagation route of the detection optical signal emitted by the auxiliary light source 104. As Figure 2 shown, when there is dirt attached to the lidar window 101, after the laser signal emitted by the VCSEL laser emitter is reflected by the target object in the environment, part of the echo signal is re-emitted by the dirt located on the lidar window 101 to the external environment and cannot be directly transmitted through the lidar window 101 to be received by the echo signal receiving unit 103. Therefore, the echo signal receiving unit 103 can only receive part of the echo energy, which will obviously affect the accuracy of the lidar's distance detection of the target object. Therefore, as Figure 2 shown, the present disclosure provides an auxiliary light source 104 in the receiving cavity. The auxiliary light source 104 can emit a detection optical signal towards the lidar window 101. In this way, when there is dirt attached to the lidar window 101, the detection optical signal can be reflected by the dirt on the lidar window, and the reflected optical signal can be received by the echo signal receiving unit 103 (such as Figure 2 the dashed arrow part in). In this way, after the echo signal receiving unit 103 converts the received reflected optical signal into an electrical signal, performs analog-to-digital conversion according to the electrical signal, and determines the real-time target image of the lidar window 101, so that the control unit can perform dirt detection on the lidar window 101 based on the target image. The above examples are only illustrative, and the present disclosure is not limited thereto.

[0075] By using the above device, by setting an auxiliary light source in the receiving cavity of the lidar (i.e., the receiving cavity where the echo signal receiving unit is located), the detection of the dirt on the lidar window can be realized, thus avoiding the problem that the dirt detection depends on the existence of the target object in the external environment. Moreover, compared with the solution of adding a CMOS or CCD optoelectronic sensor inside the lidar to realize the dirt detection without a target object, since most of the auxiliary light sources can select infrared light sources with lower costs, the disclosure can reduce the detection cost by setting an auxiliary light source to detect the dirt on the lidar window.

[0076] The following describes the specific implementation manner of the control unit 102 for detecting the dirt on the lidar window 101 based on the target image and the preset reference image.

[0077] Optionally, the control unit 102 can be configured to subtract the gray values of each pixel point of the target image and the preset reference image to obtain a dirt image; perform a histogram statistics on the dirt image to obtain a dirt detection result, and the dirt detection result includes at least one target gray value corresponding to the dirt pixel points and the number of pixel points corresponding to each target gray value.

[0078] In one implementation manner, the control unit 102 can be configured to, for each first pixel point of the target image, obtain the gray difference value between the gray value of the first pixel point of the target image and the gray value of the second pixel point on the preset reference image, where the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image; use the first pixel points with the gray difference value greater than or equal to the preset difference threshold as the third pixel points; replace the gray value of the third pixel points with the gray difference value, and set the gray values of the other pixel points except the third pixel points to the preset gray value to obtain a dirt image. Here, the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image, which can be understood that the position coordinates of the first pixel point are the same as those of the second pixel point.

[0079] As described above, the preset reference image can be a clean image pre-set when there is no dirt on the lidar window 101, that is to say, it is default that the preset reference image has no dirt. In this way, after subtracting the target image and the preset reference image pixel by pixel, for the pixel points with the gray difference value close to 0, generally it can be considered that the pixel points do not belong to the dirt pixel points, but if the gray difference value between two corresponding pixel points is large, generally it can be considered that the corresponding pixel points on the target image are the dirt pixel points.

[0080] Therefore, in the present disclosure, the third pixel point on the target image whose corresponding gray difference value is greater than or equal to the preset difference threshold can be regarded as a dirty pixel point, and the other pixel points on the target image except the third pixel point can be regarded as non-dirty pixel points. In this way, the present disclosure can replace the gray value of the third pixel point with the gray difference value, and set the gray values of the other pixel points except the third pixel point to the preset gray value to obtain a dirty image. For example, the preset gray value can be 255 (i.e., set the non-dirty image area to white).

[0081] Exemplarily, the dirty image can be obtained through the following formula:

[0082]

[0083] where B t (i, j) represents the gray value of the pixel point at the corresponding position coordinate (i, j) in the dirty image; I t (i, j) represents the gray value of the pixel point at the corresponding position coordinate (i, j) in the target image; I o (i, j) represents the gray value of the pixel point at the corresponding position coordinate (i, j) in the preset reference image; K represents the gray difference between the two pixel points, and m represents the preset difference threshold.

[0084] After obtaining the dirty image, the dirty image can be input into the control unit 102. The control unit 102 can perform histogram statistics based on the dirty image to obtain a dirty detection result. In the present disclosure, the dirty detection result can include at least one target gray value corresponding to the dirty pixel points in the preset gray value interval, and the number of pixel points corresponding to each target gray value. The preset gray value interval can be, for example, [0, 255].

[0085] Optionally, the control unit 102 can be configured to determine the occurrence probability of each gray value corresponding to the dirty image; perform normalization processing on the dirty image so as to normalize the gray value of each pixel point in the dirty image to the preset gray value interval; and then determine at least one target gray value corresponding to the dirty pixel points in the preset gray value interval and the number of pixel points corresponding to each target gray value according to the occurrence probability.

[0086] Exemplarily, the control unit 102 can perform histogram statistics based on the dirty image to obtain a dirty detection result based on the following formula:

[0087]

[0088]

[0089] Wherein, formula (2) is the statistical calculation of each gray value r in the dirty image based on formula (2) before normalizing the dirty image. k The probability of occurrence P(rk), (i, j) represents the coordinates of pixel n, M and N represent the resolution of the image; N rk Represents the gray value r k The number of pixels in the dirty image before normalization; b represents the number of bits of the dirty image (for example, b=8); S k Represents the gray value r in the preset gray value range [0,255] after the dirty image is normalized. k The number of pixels in the normalized dirty image. In this way, based on the above formulas (2) and (3), the number of pixels corresponding to each grayscale value in the preset grayscale value interval [0, 255] in the normalized dirty image can be obtained, and then the statistical result of the grayscale histogram of the dirty image can be obtained.

[0090] For example, Figure 3 is a schematic diagram of a dirt histogram curve according to an exemplary embodiment. Figure 3 As shown, except for the grayscale value of most pixels being 255 (white background), dirty pixels appear near the grayscale value of 170 (ie, the target grayscale value), and the grayscale values ​​of the dirty pixels are visible and the number is clear. This is only an example, and the present disclosure does not limit this.

[0091] Optionally, Figure 4 is based on Figure 1 The structural block diagram of a dirt detection device shown in the embodiment shown in FIG. Figure 4 As shown, the device 100 may also include a dirt processing execution unit 105 connected to the control unit, so that the control unit 102 may also be configured to send the dirt detection result to the dirt processing execution unit 105; the dirt processing execution unit 105 may be configured to perform a preset dirt cleaning operation on the lidar window film 101 according to the dirt detection result.

[0092] As described above, the dirt detection result may include at least one target gray value corresponding to the dirt pixel points within a preset gray value range, and the number of pixel points corresponding to each target gray value. In this way, for each target gray value, when the dirt processing execution unit 105 determines that the number of dirt pixel points corresponding to the target gray value is greater than or equal to a preset quantity threshold (indicating that the dirt area corresponding to the target gray value is relatively large), it may initiate a preset dirt cleaning operation. For example, the dirt processing execution unit 105 may include a driving motor and a cleaning tool (such as a cleaning rubber, a cleaning cloth, a cleaning brush, etc.), and by controlling the cleaning tool to wipe the surface of the lidar window sheet back and forth through the driving motor, the purpose of removing the dirt attached to the lidar window sheet can be achieved.

[0093] In addition, the magnitude of the target gray value in the present disclosure can ensure the degree of dirt pollution, and the number of pixel points corresponding to the target gray value can represent the dirt pollution area. Therefore, the dirt detection device provided in the present disclosure can quantitatively judge the dirt pollution degree and pollution area of the lidar window sheet. Moreover, applying the traditional digital image histogram statistical algorithm to lidar pixel-level dirt detection has low algorithm complexity and high efficiency, and the algorithm can be deployed on any processor platform; with digital image processing means, the dirt judgment accuracy is high; compared with traditional dirt detection schemes, it is not affected by the electromagnetic signals inside the lidar body and has a lower misjudgment rate.

[0094] Optionally, as Figure 4 shown, the device 100 may further include a laser emitting unit 106; in this way, the control unit 102 may also be configured to send a notification message indicating the completion of dirt cleaning to the laser emitting unit 106; the laser emitting unit 106 may be configured to emit a laser signal outward after receiving the notification message indicating the completion of dirt cleaning, so as to detect a target object in the environment based on the lidar. Since the dirt on the lidar window sheet has been cleaned, the problem of energy loss in the reception of the laser echo signal by the lidar sensor caused by the dirt on the window sheet can be avoided, and the accuracy and efficiency of target object detection can be improved.

[0095] It should also be noted that the emission power of the auxiliary light source 104 for emitting the detection optical signal is also related to the exposure rate of the reflected optical signal of the detection optical signal received by the echo signal receiving unit 103. If the echo signal receiving unit 103 is overexposed, it will inevitably affect the reception of the reflected optical signal, and further affect the dirt detection accuracy. Therefore, in the present disclosure, the control unit 102 may also be configured to obtain the current exposure rate of the echo signal receiving unit 103 and control the emission power of the auxiliary light source 104 for emitting the detection optical signal according to the exposure rate. To avoid overexposure of the echo signal receiving unit 103.

[0096] Figure 5is a flowchart of a dirt detection method shown according to an exemplary embodiment, and this method can be applied to Figure 1 the dirt detection device shown, and this device includes: a lidar window sheet, a control unit, an echo signal receiving unit and an auxiliary light source respectively connected to the control unit; wherein, the auxiliary light source and the echo signal receiving unit are arranged in the same cavity; as Figure 5 shown, this method includes the following steps:

[0097] In step S501, a detection optical signal is emitted to the lidar window sheet through the auxiliary light source.

[0098] In one implementation, the auxiliary light source can select a low-cost infrared LED device or a laser emitter. It should be noted that the wavelength of the optical signal emitted by the auxiliary light source needs to be the same as the wavelength of the laser emitted by the laser emitter of the lidar. In addition, when setting the auxiliary light source, the installation angle of the auxiliary light source needs to be arranged as a target angle, where the target angle is the angle that can make the optical detection signal emitted by the auxiliary light source cover the entire optical receiving area window sheet (usually referring to the area of the lidar window sheet directly opposite to the lidar receiving cavity).

[0099] In step S502, after the echo signal receiving unit receives the reflected optical signal of the detection optical signal reflected by the lidar window sheet, it determines the target image of the lidar window sheet according to the reflected optical signal.

[0100] Among them, the echo signal receiving unit can include a lidar sensor located in the lidar receiving cavity for receiving the echo energy emitted by the laser emitter (located in the lidar transmitting cavity).

[0101] In one implementation, the echo signal receiving unit can convert the reflected optical signal into an electrical signal; after performing analog-to-digital conversion according to the electrical signal, it determines the target image.

[0102] In step S503, the lidar window sheet is subjected to dirt detection through the control unit according to the target image and a preset reference image.

[0103] Among them, the preset reference image can include a clean image pre-set when the lidar window sheet is not dirty.

[0104] By adopting the above method, by arranging an auxiliary light source in the receiving cavity of the lidar (i.e., the receiving cavity where the echo signal receiving unit is located), the detection of the dirt on the lidar window can be realized, thus avoiding the problem that the dirt detection depends on the existence of the target object in the external environment. Moreover, compared with the dirt detection scheme without a target object by adding a CMOS or CCD optoelectronic sensor inside the lidar, since most of the auxiliary light sources can select infrared light sources with lower costs, the disclosure can reduce the detection cost by arranging the auxiliary light source to detect the dirt on the lidar window.

[0105] Figure 6 is based on Figure 5 The flowchart of a dirt detection method shown in the illustrated embodiment is as follows Figure 6 As shown, step S503 includes the following sub-steps:

[0106] In step S5031, a dirt image is obtained by subtracting the gray values of each pixel point of the target image and the preset reference image.

[0107] In this step, for each first pixel point of the target image, the gray difference value between the gray value of the first pixel point of the target image and the gray value of the second pixel point on the preset reference image is obtained, and the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image; the first pixel points with gray difference values greater than or equal to the preset difference threshold are used as the third pixel points; after replacing the gray value of the third pixel points with the gray difference value and setting the gray values of other pixel points except the third pixel points to the preset gray value, a dirt image is obtained.

[0108] Among them, the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image, which can be understood that the position coordinates of the first pixel point are the same as those of the second pixel point.

[0109] As described above, the preset reference image can be a clean image preset when there is no dirt on the lidar window 101, that is to say, it is default that the preset reference image has no dirt. In this way, after subtracting the target image and the preset reference image pixel by pixel, for the pixel points with gray difference values close to 0, generally it can be considered that the pixel points do not belong to the dirt pixel points, but if the gray difference values of two corresponding pixel points are large, generally it can be considered that the corresponding pixel points on the target image are the dirt pixel points.

[0110] Therefore, in the present disclosure, the third pixel point on the target image corresponding to a gray-scale difference greater than or equal to a preset difference threshold can be regarded as a dirty pixel point, and other pixel points on the target image except the third pixel point can be regarded as non-dirty pixel points. In this way, the present disclosure can replace the gray-scale value of the third pixel point with the gray-scale difference, and set the gray-scale values of other pixel points except the third pixel point to a preset gray-scale value to obtain a dirty image. For example, the preset gray-scale value can be 255 (i.e., set the image area that is not dirty to white).

[0111] In step S5032, perform a histogram statistics on the dirty image to obtain a dirty detection result, where the dirty detection result includes at least one target gray-scale value corresponding to the dirty pixel points and the number of pixel points corresponding to each target gray-scale value.

[0112] After obtaining the dirty image, the dirty image can be input into the control unit, and the control unit can perform a histogram statistics based on the dirty image to obtain a dirty detection result.

[0113] In this step, the control unit can perform a histogram statistics on the dirty image in the following manner to obtain a dirty detection result: for each gray-scale value corresponding to the dirty image, determine the occurrence probability of the gray-scale value in the dirty image; perform a normalization process on the dirty image so as to normalize the gray-scale value of each pixel point in the dirty image to a preset gray-scale value interval; determine at least one target gray-scale value corresponding to the dirty pixel points in the preset gray-scale value interval according to the occurrence probability, and the number of pixel points corresponding to each target gray-scale value.

[0114] In one implementation, the dirty detection device may further include: a dirty processing execution unit connected to the control unit. Figure 7 is a flowchart of a dirty detection method shown in the embodiment according to Figure 5 As shown in Figure 7 shown, the method further includes the following steps:

[0115] In step S504, send the dirty detection result to the dirty processing execution unit through the control unit.

[0116] In step S505, according to the dirty detection result, perform a preset dirty cleaning operation on the lidar window through the dirty processing execution unit.

[0117] As described above, the dirt detection result may include at least one target gray value corresponding to the dirt pixel points within a preset gray value range, and the number of pixel points corresponding to each target gray value. In this way, for each target gray value, when the dirt processing execution unit determines that the number of dirt pixel points corresponding to the target gray value is greater than or equal to a preset number threshold (indicating a relatively large dirt area corresponding to the target gray value), a preset dirt cleaning operation may be started. For example, the dirt processing execution unit may include a driving motor and a cleaning tool (such as a cleaning rubber, a cleaning cloth, a cleaning brush, etc.), and by controlling the cleaning tool to wipe the surface of the lidar window pane back and forth through the driving motor, the purpose of removing the dirt attached to the lidar window pane can be achieved.

[0118] The magnitude of the target gray value in the present disclosure can ensure the degree of dirt pollution, and the number of pixel points corresponding to the target gray value can represent the dirt pollution area. Therefore, the dirt detection device provided in the present disclosure can quantitatively judge the dirt pollution degree and pollution area of the lidar window pane. Moreover, by applying the traditional digital image histogram statistical algorithm to the lidar pixel-level dirt detection, the algorithm complexity is low, the efficiency is high, and the algorithm can be deployed on any processor platform; with digital image processing means, the dirt judgment accuracy is high; compared with the traditional dirt detection scheme, it is not affected by the electromagnetic signals inside the lidar body, and the misjudgment rate is lower.

[0119] In one implementation, the dirt detection device may further include: a laser emitting unit. Figure 8 is according to Figure 7 The flowchart of a dirt detection method shown in the illustrated embodiment, as Figure 8 shown, the method further includes the following steps:

[0120] In step S506, the control unit sends a notification message indicating the completion of dirt cleaning to the laser emitting unit; after receiving the notification message, the laser emitting unit emits a laser signal outward.

[0121] After receiving the notification message, the laser emitting unit emits a laser signal outward, which can realize the detection of the target object in the environment. Since the dirt on the lidar window pane has been cleaned, the problem of energy loss of the lidar sensor receiving the laser echo signal caused by the dirt on the window pane can be avoided, and the accuracy and efficiency of target object detection can be improved.

[0122] It should also be noted that the magnitude of the emission power of the auxiliary light source for emitting the detection optical signal is also related to the exposure rate of the reflected optical signal of the detection optical signal received by the echo signal receiving unit. If the echo signal receiving unit is overexposed, it will inevitably affect the reception of the reflected optical signal, thereby affecting the accuracy of dirt detection. Therefore, in the present disclosure, the control unit can also obtain the current exposure rate of the echo signal receiving unit and control the emission power of the auxiliary light source for emitting the detection optical signal according to the exposure rate to avoid overexposure of the echo signal receiving unit.

[0123] The present disclosure also provides a computer-readable storage medium, on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of the dirt detection method provided by the present disclosure are implemented.

[0124] Figure 9 is a structural block diagram of a lidar shown according to an exemplary embodiment, as Figure 9 shown, the lidar 900 provided by the present disclosure may include the dirt detection device 100 described above.

[0125] Figure 10 is a structural block diagram of a vehicle shown according to an exemplary embodiment, as Figure 10 shown, the vehicle 1000 provided by the present disclosure may include the lidar 900 described above.

[0126] Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and practicing the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common general knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0127] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A dirt detection device, characterized in that, it includes: a lidar window pane, a control unit, an echo signal receiving unit and an auxiliary light source respectively connected to the control unit; wherein, the auxiliary light source and the echo signal receiving unit are arranged in the same cavity; the auxiliary light source is configured to emit a detection optical signal to the lidar window pane; the echo signal receiving unit is configured to determine a target image of the lidar window pane according to the reflected optical signal after receiving the reflected optical signal of the detection optical signal reflected by the lidar window pane; the control unit is configured to perform dirt detection on the lidar window pane according to the target image and a preset reference image.

2. The device according to claim 1, characterized in that, the control unit is configured to subtract the gray values of each pixel point of the target image and the preset reference image to obtain a dirt image; perform histogram statistics on the dirt image to obtain a dirt detection result, and the dirt detection result includes at least one target gray value corresponding to the pixel points of the dirt and the number of pixel points corresponding to each target gray value.

3. The device according to claim 2, characterized in that, the control unit is configured to, for each first pixel point of the target image, obtain the gray value difference between the gray value of the first pixel point of the target image and the gray value of the second pixel point on the preset reference image, and the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image; use the first pixel point with the gray value difference greater than or equal to a preset difference threshold as the third pixel point; replace the gray value of the third pixel point with the gray value difference, and set the gray values of other pixel points except the third pixel point to a preset gray value to obtain the dirt image.

4. The device according to claim 2, characterized in that, the control unit is configured to determine the occurrence probability of each gray value corresponding to the dirt image in the dirt image; perform normalization processing on the dirt image so as to normalize the gray value of each pixel point in the dirt image to a preset gray value interval; determine at least one target gray value corresponding to the pixel points of the dirt in the preset gray value interval according to the occurrence probability, and the number of pixel points corresponding to each target gray value.

5. The device according to claim 1, characterized in that, the echo signal receiving unit is configured to convert the reflected optical signal into an electrical signal, and then perform analog-to-digital conversion according to the electrical signal to determine the target image.

6. The device according to claim 1, characterized in that, the control unit is further configured to obtain the current exposure rate of the echo signal receiving unit and control the emission power of the detection optical signal emitted by the auxiliary light source according to the exposure rate.

7. The device according to any one of claims 1-6, characterized in that, the device further includes: a dirt treatment execution unit connected to the control unit; The control unit is further configured to send the dirt detection result to the dirt treatment execution unit; The dirt treatment execution unit is configured to perform a preset dirt cleaning operation on the lidar window according to the dirt detection result.

8. The device according to claim 7, wherein, The device further includes: a laser emission unit; The control unit is further configured to send a notification message indicating the completion of dirt cleaning to the laser emission unit; The laser emission unit is configured to emit a laser signal outward after receiving the notification message.

9. A dirt detection method, wherein, Applied to a dirt detection device, the device includes: a lidar window, a control unit, an echo signal receiving unit and an auxiliary light source respectively connected to the control unit; wherein, the auxiliary light source and the echo signal receiving unit are arranged in the same cavity; the method includes: Emitting a detection optical signal to the lidar window through the auxiliary light source; After the echo signal receiving unit receives the reflected optical signal of the detection optical signal reflected by the lidar window, determining a target image of the lidar window according to the reflected optical signal; Performing dirt detection on the lidar window through the control unit according to the target image and a preset reference image.

10. The method according to claim 9, wherein, The performing dirt detection on the lidar window through the control unit according to the target image and a preset reference image includes: Subtracting the gray values of each pixel point of the target image and the preset reference image to obtain a dirt image; Performing a histogram statistics on the dirt image to obtain a dirt detection result, the dirt detection result including at least one target gray value corresponding to the pixel points of the dirt and the number of pixel points corresponding to each target gray value.

11. The method according to claim 10, wherein, The subtracting the gray values of each pixel point of the target image and the preset reference image to obtain a dirt image includes: For each first pixel point of the target image, obtaining the gray difference value between the gray value of the first pixel point of the target image and the gray value of the second pixel point on the preset reference image, and the second pixel point is the pixel point corresponding to the first pixel point on the preset reference image; Taking the first pixel points whose gray difference value is greater than or equal to a preset difference threshold as third pixel points; Replacing the gray value of the third pixel point with the gray difference value, and setting the gray values of other pixel points except the third pixel point to a preset gray value to obtain the dirt image.

12. The method according to claim 10, wherein, The performing a histogram statistics on the dirt image to obtain a dirt detection result includes: For each gray value corresponding to the dirt image, determining the occurrence probability of the gray value in the dirt image; Performing a normalization process on the dirt image so as to normalize the gray value of each pixel point in the dirt image to a preset gray value interval; Determine at least one of the target gray values corresponding to the dirty pixel points in the preset gray value interval according to the occurrence probability, and the number of pixel points corresponding to each of the target gray values.

13. The method according to claim 9, wherein, the determining the target image of the lidar window sheet according to the reflected light signal includes: converting the reflected light signal into an electrical signal; after performing analog-to-digital conversion according to the electrical signal, determining the target image.

14. The method according to claim 9, wherein, the method further includes: acquiring, by the control unit, the current exposure rate of the echo signal receiving unit, and controlling the emission power of the detection light signal emitted by the auxiliary light source according to the exposure rate.

15. The method according to any one of claims 9-14, wherein, the device further includes: a dirt treatment execution unit connected to the control unit; the method further includes: sending, by the control unit, a dirt detection result to the dirt treatment execution unit; performing a preset dirt cleaning operation on the lidar window sheet by the dirt treatment execution unit according to the dirt detection result.

16. The method according to claim 15, wherein, the device further includes: a laser emission unit; the method further includes: sending, by the control unit, a notification message indicating that the dirt cleaning is completed to the laser emission unit; after receiving the notification message, the laser emission unit emits a laser signal outward.

17. A computer-readable storage medium, on which computer program instructions are stored, wherein, when the program instructions are executed by a processor, the steps of the method according to any one of claims 9-16 are implemented.

18. A lidar, wherein, it includes the dirt detection device according to any one of claims 1-8.

19. A vehicle, wherein, it includes the lidar according to claim 18.

Images