A target identification method, device, apparatus and storage medium

Abstract

Embodiments of the present application disclose a target identification method, device, setting and storage medium, wherein the method comprises: acquiring lane attribute information of a target in a road monitoring range; determining other feature information of the target matched with the acquired lane attribute information; and identifying the type of the target by using the lane attribute information and the other feature information. Embodiments of the present application can identify the type of the target by using the lane attribute of the target and combining the other feature information corresponding to the lane attribute, so that the target can be accurately detected in the intelligent traffic scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traffic target detection, and in particular to a target recognition method, device, equipment and storage medium. BACKGROUND

[0002] In the city intelligent traffic, the camera is mostly used to recognize the target at the intersection, but in the rain and snow weather, the long-distance lane target behind the zebra crossing, and the limited complex scene such as viaduct and underground passage, it is difficult to play a role. At present, the target recognition according to the radar signal is low in detection accuracy because of less available feature information. SUMMARY

[0003] Therefore, the embodiments of the present application provide a target recognition method, device, equipment and storage medium, so as to accurately detect the target in the intelligent traffic scene.

[0004] In a first aspect, the embodiments of the present application provide a target recognition method, comprising:

[0005] acquiring lane attribute information of a target in a road monitoring range;

[0006] determining other feature information of the target matched with the acquired lane attribute information;

[0007] using the lane attribute information and the other feature information to perform type recognition on the target.

[0008] Further, the acquiring of the lane attribute information of the target in the road monitoring range comprises:

[0009] determining a road region where the target is located according to a trajectory of the target in the road monitoring range;

[0010] finding a mapping relationship between a plurality of road regions and a plurality of lane attributes pre-divided, to obtain a lane attribute corresponding to the road region where the target is located as the lane attribute information of the target.

[0011] Further, the determining of the road region where the target is located according to the trajectory of the target in the road monitoring range comprises:

[0012] determining the road region where the target is located according to a position where the target first appears in the road monitoring range.

[0013] Further, the determining of the other feature information of the target matched with the acquired lane attribute information comprises:

[0014] if the acquired lane attribute information is a motor lane, the other feature information of the target matched with the lane attribute information comprises a length of the target;

[0015] If the obtained lane attribute information is for a non-motorized vehicle lane, then other characteristic information of the matching target is determined, including the target's speed and / or radar cross section.

[0016] If the obtained lane attribute information is a zebra crossing, then other characteristic information of the matching target is determined, including: the target's side slip angle, and the target's speed and / or radar cross section.

[0017] Furthermore, using the lane attribute information and other feature information, target type identification is performed, including:

[0018] When the target's lane attribute information is a motor vehicle lane, the target is identified as a small vehicle, medium vehicle, or large vehicle according to its length.

[0019] When the target's lane attribute information is non-motorized vehicle lane, the target is identified as a person or non-motorized vehicle according to the target's speed and / or radar cross section;

[0020] When the target's lane attribute information is zebra crossing, determine whether the target is crossing the road based on the target's side slip angle; if so, identify the target as a person or non-motorized vehicle according to the target's speed and / or radar cross section.

[0021] Furthermore, after identifying the target as a small vehicle, medium vehicle, or large vehicle based on its length, the process also includes:

[0022] When the identified vehicle type changes from a large or medium-sized vehicle to a small vehicle, or from a large vehicle to a medium-sized vehicle, it is determined that an error has occurred in target identification.

[0023] Furthermore, the method also includes:

[0024] Estimate the length of the target;

[0025] After performing Kalman filtering on each video frame containing the target, the target length is updated when the target is traveling straight on the road.

[0026] Secondly, embodiments of the present invention provide a target recognition device, comprising:

[0027] Lane attribute information acquisition unit, used to acquire lane attribute information of targets within the road monitoring range;

[0028] The feature information determination unit is used to determine other feature information of the target that matches the acquired lane attribute information;

[0029] The identification unit is used to identify the type of the target by using the lane attribute information obtained by the lane attribute information acquisition unit and other feature information of the target determined by the feature information determination unit.

[0030] Thirdly, embodiments of the present invention provide an electronic device, including: a housing, a processor, a memory, a circuit board, and a power supply circuit, wherein the circuit board is disposed inside the space enclosed by the housing, and the processor and the memory are disposed on the circuit board; the power supply circuit is used to supply power to various circuits or devices of the above-mentioned electronic device; the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, for executing the target recognition method described in the first aspect above.

[0031] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing one or more programs, which can be executed by one or more central processing units to implement the target recognition method described in the first aspect.

[0032] The technical solution provided by this invention utilizes the lane attributes of the target and combines them with other feature information corresponding to the lane attributes to identify the type of the target, which can accurately detect the target in intelligent transportation scenarios. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 A flowchart of a target recognition method provided in an embodiment of the present invention;

[0035] Figure 2 This is a schematic diagram of the structure of a target recognition device provided in an embodiment of the present invention;

[0036] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0037] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0038] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0039] This invention provides a target recognition method, which can be executed by a corresponding target recognition device, which can be integrated into an intersection monitoring system equipped with cameras and radar. See also... Figure 1 The method specifically includes the following steps 101-103.

[0040] Step 101: Obtain lane attribute information of the target within the road monitoring range.

[0041] In this step, video footage from the road monitoring area can be pre-captured using cameras, and then the video frames can be analyzed to initially identify targets and their trajectories. Alternatively, targets and their trajectories within the road monitoring area can be pre-identified using radar illumination. Radar is often installed opposite the intersection to be monitored; for example, a radar installed on the right side of an intersection monitors targets within the zebra crossing and lane area on the left. This initial target identification process is existing technology and will not be elaborated upon here.

[0042] The lane attribute information of a target describes the lane in which the target is located. Lanes can be categorized as: motor vehicle lanes, non-motor vehicle lanes, median strips, and crosswalks. Typically, motor vehicle lanes can be further divided into: straight lanes, right-turn lanes, left-turn lanes, and left-turn lanes (including lanes where a turn is expected). For example, obtaining the lane attribute information of a target within the road monitoring range may include:

[0043] Based on the trajectory of the target within the road monitoring range, determine the road area where the target is located;

[0044] Find the mapping relationship between multiple pre-divided road regions and multiple lane attributes to obtain the lane attributes corresponding to the road region where the target is located, and use them as the lane attribute information of the target.

[0045] Preferably, the road area where the target is located is determined according to the location where the target first appears within the road monitoring range.

[0046] Step 102: Determine other feature information of the target that matches the acquired lane attribute information.

[0047] Step 103: Use the lane attribute information and other feature information to identify the type of the target.

[0048] In specific implementation, if the obtained lane attribute information is for a motor vehicle lane, other characteristic information of the matching target is determined, including the target's length. The target is then identified as a small vehicle, medium-sized vehicle, or large vehicle based on its length. For example, when the length of a target in a motor vehicle lane is less than a preset first threshold, the target is identified as a small vehicle; when the length of a target in a motor vehicle lane is greater than a preset second threshold, the target is identified as a large vehicle; and when the length of a target in a motor vehicle lane is between the preset first and second thresholds, the target is identified as a medium-sized vehicle. The first and second thresholds can be set by those skilled in the art based on experience; typically, the first threshold is 6 meters and the second threshold is 13 meters.

[0049] The calculation of the target length can be achieved using existing technology. In this embodiment of the invention, the centroid of the target can be taken as the midpoint of all the points on the target, and the position of the target can be extracted. For the aspect ratio of the target, the x-coordinates x1 and x2 of the two points with the smallest and largest horizontal axes, and the y1 and y2 of the two points with the smallest and largest vertical axes, can be taken. Based on the two vertices (x1, y1) and (x2, y2) of the rectangle, the length and width of the target can be estimated respectively.

[0050] If the acquired lane attribute information indicates a non-motorized vehicle lane, then other characteristic information of the matching target is determined, including the target's speed and / or RCS (Radar Cross Section). Based on the target's speed and / or RCS, the target is identified as either a pedestrian or a non-motorized vehicle. Since pedestrians and non-motorized vehicles differ significantly in speed and RCS, in non-motorized vehicle lane areas, targets generated in this area can be classified as pedestrians or non-motorized vehicles based on the difference in their speed and RCS.

[0051] If the obtained lane attribute information indicates a zebra crossing, other characteristic information for matching targets includes: the target's side slip angle, and the target's speed and / or radar cross section (RCS). The side slip angle is the angle between the longitudinal velocity and the resultant velocity. The side slip angle is used to distinguish between straight-moving and crossing targets. For example, the side slip angle can be used to determine if a target is crossing; if so, the target is identified as a pedestrian or non-motorized vehicle based on its speed and / or RCS. In zebra crossing areas, crossing targets can be first screened based on the side slip angle, and then classified as pedestrians or non-motorized vehicles based on the differences in speed and RCS between pedestrians and non-motorized vehicles. Straight-moving targets in zebra crossing areas are not considered until they enter the lane area for target type identification.

[0052] In a preferred embodiment, after identifying a target as a small vehicle, medium vehicle, or large vehicle based on its length, the method further includes: determining that a target identification error has occurred when the identified vehicle type changes from large or medium vehicle to small vehicle, or from large vehicle to medium vehicle. For example, the target's vehicle type is initialized as small vehicle. In this embodiment, all targets initially appearing in the traffic lane area are defaulted to small vehicles, and their length is accumulated over multiple frames. When the length reaches the classification threshold for medium or large vehicles, they are classified into the corresponding type. When updating the vehicle type, it can only change in the order of small -> medium -> large vehicle, and cannot change from large vehicle to small vehicle.

[0053] Furthermore, based on the above technical solutions, the target recognition method provided in this embodiment of the invention further includes: estimating the length of the target; updating the length of the target after performing Kalman filtering on each video frame containing the target, and when the target is traveling straight on the road. The length of the target is not updated when the target is crossing the road.

[0054] According to the different lane functions, each target is assigned a corresponding lane attribute when the trajectory of each target is generated. Combined with the target's speed, RCS, length, side slip angle and other feature information, the target is classified using a classification method (such as decision tree) to obtain 5 types of human and vehicle recognition results.

[0055] Furthermore, embodiments of the present invention also provide a target recognition device, see [link to previous document]. Figure 2 The device includes:

[0056] Lane attribute information acquisition unit 201 is used to acquire lane attribute information of targets within the road monitoring range;

[0057] The feature information determination unit 202 is used to determine other feature information of the target that matches the acquired lane attribute information;

[0058] The identification unit 203 is used to identify the type of the target by using the lane attribute information obtained by the lane attribute information acquisition unit and other feature information of the target determined by the feature information determination unit.

[0059] For example, the lane attribute information acquisition unit 201 is used to acquire lane attribute information of targets within the road monitoring range, including:

[0060] Based on the trajectory of the target within the road monitoring range, determine the road area where the target is located;

[0061] Find the mapping relationship between multiple pre-divided road regions and multiple lane attributes to obtain the lane attributes corresponding to the road region where the target is located, and use them as the lane attribute information of the target.

[0062] For example, the lane attribute information acquisition unit 201 is used to determine the road area where the target is located based on the trajectory of the target within the road monitoring range, including:

[0063] Determine the road area where the target is located based on its first appearance within the road monitoring range.

[0064] In the example row, the feature information determination unit 202 is used to determine other feature information of the target that matches the acquired lane attribute information, including:

[0065] If the obtained lane attribute information is a motor vehicle lane, then determine other characteristic information of the matching target, including the length of the target;

[0066] If the obtained lane attribute information is a motor vehicle lane, then other characteristic information of the matching target is determined, including the target's speed and / or radar cross section.

[0067] If the obtained lane attribute information is a zebra crossing, then other characteristic information of the matching target is determined, including: the target's side slip angle, and the target's speed and / or radar cross section.

[0068] For example, the identification unit 203 is used to identify the type of the target using the lane attribute information and the other feature information, including:

[0069] When the target's lane attribute information is a motor vehicle lane, the target is identified as a small vehicle, medium vehicle, or large vehicle according to its length.

[0070] When the target's lane attribute information is non-motorized vehicle lane, the target is identified as a person or non-motorized vehicle according to the target's speed and / or radar cross section;

[0071] When the target's lane attribute information is zebra crossing, determine whether the target is crossing the road based on the target's side slip angle; if so, identify the target as a person or non-motorized vehicle according to the target's speed and / or radar cross section.

[0072] For example, the target recognition device provided in this embodiment of the invention further includes an error judgment unit 204, used for:

[0073] After the identification unit 203 identifies the target as a small vehicle, medium vehicle, or large vehicle according to the length of the target, if the identification unit 203 identifies that the vehicle type of the target changes from a large vehicle or medium vehicle to a small vehicle, or a large vehicle to a medium vehicle, it determines that the target identification has been incorrect.

[0074] For example, the target recognition device provided in this embodiment of the invention further includes a target length estimation unit 200, used for: estimating the length of the target; and updating the length of the target after performing Kalman filtering on each video frame containing the target and when the target is traveling straight on the road.

[0075] The target identification device provided in this embodiment of the invention belongs to the same inventive concept as the target method in the foregoing embodiments. Technical details not described in this embodiment can be found in the relevant descriptions in the foregoing embodiments, and will not be repeated here.

[0076] Furthermore, embodiments of the present invention also provide an electronic device, see [link to previous document]. Figure 3 The electronic device may include: a housing 31, a processor 32, a memory 33, a circuit board 34, and a power supply circuit 35. The circuit board 34 is disposed within the space enclosed by the housing 31, and the processor 32 and memory 33 are mounted on the circuit board 34. The power supply circuit 35 supplies power to the various circuits or devices of the electronic device. The memory 33 stores executable program code. The processor 32 reads the executable program code stored in the memory 33 to run a program corresponding to the executable program code, thereby executing the target recognition method described in any of the foregoing embodiments. For details on the specific execution process of the processor 32 of the above steps and the steps further executed by the processor 32 through running the executable program code, please refer to this invention. Figure 1 The description of the illustrated embodiments will not be repeated here.

[0077] Finally, this embodiment of the invention also provides a computer-readable storage medium storing one or more programs, which can be executed by one or more central processing units to implement the target recognition method described in this embodiment of the invention.

[0078] In summary, the technical solution provided by the embodiments of the present invention utilizes the lane attributes of the target and combines them with other feature information corresponding to the lane attributes to identify the type of the target, which can accurately detect the target in intelligent transportation scenarios.

[0079] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0080] In this embodiment of the invention, the term "and / or" describes the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following associated objects have an "or" relationship.

[0081] The various embodiments in this specification are described in a related manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0082] In particular, the device embodiment is basically similar to the method embodiment, so the description is relatively simple. For relevant details, please refer to the description of the method embodiment.

[0083] For ease of description, the above apparatus is described by dividing it into various functional units / modules. Of course, in implementing this invention, the functions of each unit / module can be implemented in one or more software and / or hardware.

[0084] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0085] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A target recognition method, characterized in that, The method includes: Obtain lane attribute information of targets within the road monitoring range; Identify other feature information of the target that matches the acquired lane attribute information; Using the lane attribute information and other feature information, target type identification is performed; wherein, determining other feature information of the target that matches the acquired lane attribute information includes: If the obtained lane attribute information is a motor vehicle lane, then determine other characteristic information of the matching target, including the length of the target; If the obtained lane attribute information is for a non-motorized vehicle lane, then other characteristic information of the matching target is determined, including the target's speed and / or radar cross section. If the obtained lane attribute information is a zebra crossing, then other characteristic information of the matching target is determined to include: the target's sideslip angle, and the target's speed and / or radar cross section. Using the lane attribute information and other feature information, target type identification is performed, including: When the target's lane attribute information is a motor vehicle lane, the target is identified as a small vehicle, medium vehicle, or large vehicle according to its length. When the target's lane attribute information is non-motorized vehicle lane, the target is identified as a person or non-motorized vehicle according to the target's speed and / or radar cross section; When the target's lane attribute information is zebra crossing, determine whether the target is crossing the road based on the target's side slip angle; if so, identify the target as a person or non-motorized vehicle according to the target's speed and / or radar cross section.

2. The method according to claim 1, characterized in that, Obtain lane attribute information of targets within the road monitoring range, including: Based on the trajectory of the target within the road monitoring range, determine the road area where the target is located; Find the mapping relationship between multiple pre-divided road regions and multiple lane attributes to obtain the lane attributes corresponding to the road region where the target is located, and use them as the lane attribute information of the target.

3. The method according to claim 2, characterized in that, Based on the trajectory of the target within the road monitoring range, determine the road area where the target is located, including: Determine the road area where the target is located based on its first appearance within the road monitoring range.

4. The method according to claim 3, characterized in that, After identifying the target as a small vehicle, medium vehicle, or large vehicle based on its length, the process also includes: When the identified vehicle type changes from a large or medium-sized vehicle to a small vehicle, or from a large vehicle to a medium-sized vehicle, it is determined that an error has occurred in target identification.

5. The method according to any one of claims 2-4, characterized in that, The method further includes: Estimate the length of the target; After performing Kalman filtering on each video frame containing the target, the target length is updated when the target is traveling straight on the road.

6. A target recognition device, characterized in that, The device includes: Lane attribute information acquisition unit, used to acquire lane attribute information of targets within the road monitoring range; The feature information determination unit is used to determine other feature information of the target that matches the acquired lane attribute information; The identification unit is used to identify the type of the target by using the lane attribute information obtained by the lane attribute information acquisition unit and other feature information of the target determined by the feature information determination unit, wherein the feature information determination unit is specifically used for: If the obtained lane attribute information is a motor vehicle lane, then determine other characteristic information of the matching target, including the length of the target; If the obtained lane attribute information is for a non-motorized vehicle lane, then other characteristic information of the matching target is determined, including the target's speed and / or radar cross section. If the obtained lane attribute information is a zebra crossing, then other characteristic information of the matching target is determined to include: the target's sideslip angle, and the target's speed and / or radar cross section. The identification unit is specifically used for: When the target's lane attribute information is a motor vehicle lane, the target is identified as a small vehicle, medium vehicle, or large vehicle according to its length. When the target's lane attribute information is non-motorized vehicle lane, the target is identified as a person or non-motorized vehicle according to the target's speed and / or radar cross section; When the target's lane attribute information is zebra crossing, determine whether the target is crossing the road based on the target's side slip angle; if so, identify the target as a person or non-motorized vehicle according to the target's speed and / or radar cross section.

7. An electronic device, characterized in that, The electronic device includes: a housing, a processor, a memory, a circuit board, and a power supply circuit, wherein the circuit board is disposed inside the space enclosed by the housing, and the processor and the memory are disposed on the circuit board; the power supply circuit is used to supply power to various circuits or devices of the electronic device; the memory is used to store executable program code; the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, for executing the method described in any one of claims 1-5 above.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs that can be executed by one or more secure processors to implement the method described in any one of claims 1-5.

Images