Method for automatically setting region of interest, device for automatically setting region of interest, and traffic volume measurement system

Abstract

Embodiments of the present disclosure relate to a method for automatically setting a region of interest, a device for automatically setting a region of interest, and a traffic volume measurement system, which may: on the basis of image data obtained by photographing a road near a vehicle travel direction change zone through a camera, generate vehicle movement path data for vehicles driving on the road as a feature map for the road; set N lanes (N is a natural number of 1 or more) included in the road on the basis of the feature map; and set a region of interest for each of the N lanes on the basis of the feature map.

Description

Method for automatic setting of interest area, device for automatic setting of interest area, and traffic volume measurement system

[0001] The embodiments of the present disclosure relate to a method for automatically setting a region of interest, an apparatus for automatically setting a region of interest, and a traffic volume measurement system.

[0002] Currently, in the transportation sector, traffic volume measurement systems based on video are being developed. For traffic-related processing utilizing measurement results to be effectively and appropriately carried out, the system must measure traffic volume accurately and quickly. However, conventional traffic volume measurement systems face difficulties in accurately and rapidly measuring traffic volume due to various reasons, such as the road environment where measurement is required.

[0003] Embodiments of the present disclosure can provide a method for automatically setting an area of ​​interest and an apparatus for automatically setting an area of ​​interest that can accurately and quickly automatically set an area of ​​interest for traffic volume measurement.

[0004] Embodiments of the present disclosure may provide a method for automatically setting an area of ​​interest and an apparatus for automatically setting an area of ​​interest that can automatically reset the area of ​​interest when the camera's field of view changes.

[0005] The embodiments of the present disclosure can provide a traffic volume measurement system capable of accurately and quickly measuring traffic volume by accurately and quickly setting the area of ​​interest.

[0006] An automatic interest region setting device according to embodiments of the present disclosure may include a feature map generating unit that generates vehicle movement path data for vehicles traveling on the road as a feature map for the road based on image data of a road near a vehicle driving direction change zone captured through a camera, a lane setting unit that sets N lanes (N is a natural number greater than or equal to 1) included in the road based on the feature map, and an interest region setting unit that sets an interest region for each of the N lanes based on the feature map.

[0007] A traffic volume measurement system according to embodiments of the present disclosure may include an automatic area of ​​interest setting device that sets an area of ​​interest for each of N (N is a natural number greater than or equal to 1) lanes included in the road based on image data captured of a road near a vehicle driving direction change zone through a camera, and a traffic volume measurement device that measures the traffic volume for each of the areas of interest for each of the N lanes. The automatic area of ​​interest setting device may include a feature map generation unit that generates vehicle movement path data for vehicles driving on the road as a feature map for the road based on image data, a lane setting unit that sets N (N is a natural number greater than or equal to 1) lanes included in the road based on the feature map, and an area of ​​interest setting unit that sets an area of ​​interest for each of the N lanes based on the feature map.

[0008] A method for automatically setting a region of interest according to embodiments of the present disclosure may include the steps of: acquiring vehicle movement path data for vehicles traveling on a road based on image data of a road near a vehicle driving direction change zone captured through a camera, and generating a feature map for the road based on the vehicle movement path data; setting N lanes (N is a natural number greater than or equal to 1) included in the road based on the feature map; and setting a region of interest for each of the N lanes based on the feature map.

[0009] According to embodiments of the present disclosure, a method for automatically setting an area of ​​interest and an apparatus for automatically setting an area of ​​interest can be provided, which can accurately and quickly automatically set an area of ​​interest for measuring traffic volume.

[0010] According to embodiments of the present disclosure, a method for automatically setting an area of ​​interest and an apparatus for automatically setting an area of ​​interest can be provided, which can automatically reset the area of ​​interest when the angle of view of a camera changes.

[0011] According to embodiments of the present disclosure, a traffic volume measurement system can be provided that can accurately and quickly measure traffic volume by accurately and quickly setting an area of ​​interest.

[0012] FIG. 1 is a block diagram of a traffic volume measurement system according to embodiments of the present disclosure.

[0013] FIG. 2 shows the installation environment of a traffic volume measurement system according to embodiments of the present disclosure.

[0014] FIG. 3 is a block diagram of an automatic region of interest setting device according to embodiments of the present disclosure.

[0015] FIG. 4 shows an original image for setting a region of interest according to embodiments of the present disclosure.

[0016] FIGS. 5 and 6 are drawings illustrating the process of generating a feature map for setting a region of interest according to embodiments of the present disclosure.

[0017] FIGS. 7 and 8 are drawings for illustrating a car recognition process for setting a region of interest according to embodiments of the present disclosure.

[0018] FIG. 9 shows the result of setting the region of interest according to embodiments of the present disclosure.

[0019] FIG. 10 is a flowchart of a method for automatically setting a region of interest according to embodiments of the present disclosure.

[0020] The automatic region of interest setting device according to the present disclosure may be a device that automatically sets a region of interest based on data.

[0021] The automatic region of interest setting device according to the present disclosure can generate a feature map based on the position and direction of a vehicle, distinguish at least one lane on a road based on the feature map, and set a region of interest for each distinguished lane.

[0022] In the present disclosure, “lane” refers to a part of a road (road) separated by lanes (lane marking lines) for vehicles to travel in a designated line, and may also be referred to as “lane.”

[0023] The automatic interest region setting device according to the present disclosure can set the arrival direction (hereinafter also referred to as the arrival interest region) to which a vehicle reaches according to the direction of movement of the vehicle for each lane through data clustering based on vehicle movement path data.

[0024] The automatic region of interest setting device according to the present disclosure can distinguish vehicles using deep learning and obtain information about the region of interest using movement data (vehicle movement path data) of an object corresponding to a vehicle.

[0025] According to the present disclosure, for measuring traffic volume, a camera is installed to look at a vehicle driving direction change area, such as an intersection or a U-turn section, and object distinction of a vehicle in an image captured by the camera can be distinguished using deep learning.

[0026] The automatic region of interest setting device according to the present disclosure can configure movement path data for vehicles classified by deep learning in the form of a feature map. Here, the data for the feature map may include information regarding the position and direction of movement of each object (vehicle) over time.

[0027] The automatic region of interest setting device according to the present disclosure can recognize a lane in the direction of a measurement target using data from a characteristic map. The lane recognition method is as follows.

[0028] The automatic region of interest setting device according to the present disclosure can distinguish lanes through continuity and clustering with respect to the vehicle's travel path.

[0029] The automatic setting device for a region of interest according to the present disclosure can set the vehicle movement path (lane) closest to the boundary (corresponding to the center line on a road) of two groups of vehicles traveling in different opposite directions as the first lane, through clustering (grouping) of vehicles traveling in the same direction with respect to the setting of the first lane.

[0030] The automatic setting device for a region of interest according to the present disclosure can set lanes by sequentially increasing lane numbers in a direction away from the first lane.

[0031] The automatic area of ​​interest setting device according to the present disclosure can set a stop line by limiting the area where two or more vehicle driving directions overlap in a pre-set lane(s).

[0032] The automatic interest area setting device according to the present disclosure can obtain direction information by distinguishing the path (location) and direction of movement of a vehicle moving from the stop line after setting the lane and the stop line.

[0033] According to the automatic region of interest setting device of the present disclosure, the region of interest for a direction can be classified according to the following rules.

[0034] The automatic setting device for the region of interest according to the present disclosure can be set as the extinction point of vehicles moving in the opposite direction in the case of a U-turn.

[0035] The automatic region of interest setting device according to the present disclosure can measure how many directions are distinguished through clustering according to the direction of travel of each vehicle in the case of left turn, straight, and right turn.

[0036] For example, the automatic region of interest setting device according to the present disclosure can distinguish between left turn, straight, and right turn based on the number of measured directions. The automatic region of interest setting device according to the present disclosure can determine one direction as straight. In the case of two directions, the automatic region of interest setting device according to the present disclosure can distinguish between left turn and right turn based on the remaining angle of movement relative to the straight direction. In the case of three directions, the automatic region of interest setting device according to the present disclosure can distinguish between left turn and right turn based on the angle of movement relative to the straight direction.

[0037] By automatically setting the region of interest according to the method for automatically setting the region of interest executed by the device for automatically setting the region of interest according to the above-described disclosure, a significant amount of time and effort required for setting the region of interest can be saved.

[0038] In addition, according to the method for automatically setting a region of interest according to the present disclosure described above, the region of interest is set by viewing data over a certain period, thereby improving upon the existing inconvenience of having to continuously observe and modify it.

[0039] In addition, according to the method for automatically setting a region of interest according to the present disclosure described above, it is possible to respond immediately and quickly when the angle of view of the camera changes due to special factors such as a change in the angle of view caused by replacement due to camera failure or external environment.

[0040] Hereinafter, some embodiments of the present disclosure will be described in detail with reference to the exemplary drawings. In assigning reference numerals to the components of each drawing, the same components may have the same reference numeral as much as possible, even if they are shown in different drawings. Furthermore, in describing the present disclosure, if it is determined that a detailed description of related known components or functions may obscure the essence of the present disclosure, such detailed description may be omitted. Where terms such as "comprising," "having," or "consisting of" are used in this specification, other parts may be added unless "only" is used. Where a component is expressed in the singular, it may include a plural unless there is a special explicit description otherwise.

[0041] Additionally, terms such as first, second, A, B, (a), (b), etc., may be used to describe the components of the present disclosure. These terms are used merely to distinguish the components from other components, and the nature, order, sequence, or number of the components are not limited by such terms.

[0042] In describing the positional relationship of components, where it is stated that two or more components are "connected," "combined," or "joined," it should be understood that while the two or more components may be directly "connected," "combined," or "joined," they may also be "connected," "combined," or "joined" with other components "intervened." Here, the other components may be included in one or more of the two or more components that are "connected," "combined," or "joined" with one another.

[0043] In describing the temporal flow relationship regarding components, methods of operation, or methods of production, for example, when the temporal or sequential relationship is described using "after," "following," "next," or "before," it may include cases where the relationship is not continuous unless "immediately" or "directly" is used.

[0044] Meanwhile, where numerical values ​​or corresponding information regarding a component (e.g., levels, etc.) are mentioned, even without separate explicit notation, the numerical values ​​or corresponding information may be interpreted as including a range of error that may occur due to various factors (e.g., process factors, internal or external shocks, noise, etc.).

[0045] Hereinafter, various embodiments of the present disclosure will be described in detail with reference to the attached drawings.

[0046] FIG. 1 is a block diagram of a traffic volume measurement system (100) according to embodiments of the present disclosure. FIG. 2 shows the installation environment of a traffic volume measurement system (100) according to embodiments of the present disclosure.

[0047] Referring to FIG. 1, a traffic volume measuring system (100) according to embodiments of the present disclosure may include a camera (110) that photographs a road near a vehicle driving direction change area (210), an automatic area of ​​interest setting device (120) that sets an area of ​​interest for each of N lanes (N is a natural number greater than or equal to 1) included in the road based on image data captured through the camera (110), and a traffic volume measuring device (130) that measures the traffic volume for the area of ​​interest for each of the N lanes.

[0048] Referring to FIG. 2, the camera (110) may be installed on a road structure (200) near a vehicle driving direction change area (210). For example, the vehicle driving direction change area (210) may be an intersection or a U-turn area.

[0049] Referring to FIG. 2, for example, the area of ​​interest automatic setting device (120) may be installed on a road structure (200) on which the camera (110) is installed, or on a road structure other than the road structure (200). For another example, the area of ​​interest automatic setting device (120) may be a traffic control center server that receives image data from the camera (110) or may be included in the traffic control center server.

[0050] Referring to FIG. 2, the traffic volume measuring device (130) may be a traffic control center server or may be included in a traffic control center server.

[0051] According to the traffic volume measurement system (100) according to the embodiments of the present disclosure, the region of interest for each of the N lanes may be a region for measuring traffic volume.

[0052] According to the traffic volume measurement system (100) according to embodiments of the present disclosure, a region of interest for each of N lanes may include a previous region of interest in the section prior to a vehicle traveling in the lane reaching a vehicle driving direction change zone (210), and a destination region of interest to be reached after a vehicle traveling in the lane passes through the vehicle driving direction change zone (210). For example, the destination region of interest may include at least one sub-region. At least one sub-region may be a region corresponding to the vehicle driving direction.

[0053] FIG. 3 is a block diagram of an automatic region of interest setting device (120) according to embodiments of the present disclosure.

[0054] Referring to FIG. 3, an automatic region of interest setting device (120) according to embodiments of the present disclosure may include a characteristic map generating unit (320), a lane setting unit (330), and a region of interest setting unit (340).

[0055] The feature map generation unit (320) can generate vehicle movement path data for vehicles traveling on the road as a feature map of the road based on image data of the road near the vehicle driving direction change area (210) captured through the camera (110).

[0056] The lane setting unit (330) can set N lanes (N is a natural number greater than or equal to 1) included in the road based on a characteristic map corresponding to vehicle movement path data for vehicles driving on the road.

[0057] The region of interest setting unit (340) can set a region of interest for each of the N lanes based on a characteristic map.

[0058] Referring to FIG. 3, the region of interest automatic setting device (120) according to embodiments of the present disclosure may further include a deep learning-based vehicle recognition unit (310) that recognizes a vehicle driving on a road in image data through deep learning.

[0059] The feature map generation unit (320) can obtain vehicle movement path data for vehicles driving on the road based on the vehicle recognition results of the deep learning-based vehicle recognition unit (310).

[0060] Referring to FIG. 3, the automatic interest region setting device (120) according to embodiments of the present disclosure may further include an interest region resetting control unit (350) that outputs a command signal for resetting the interest region for each of N lanes when interest region resetting event information is input.

[0061] For example, the interest area reset event information may include at least one of replacement information of the camera (110), information on changing the installation location of the camera (110), information on changing the viewing angle of the camera (110), and information on changing the settings of the camera (110).

[0062] As described above, the automatic setting device (120) for a region of interest according to the embodiments of the present disclosure can set the region of interest accurately and quickly. Accordingly, the traffic volume measurement system (100) according to the embodiments of the present disclosure can measure the traffic volume accurately and quickly.

[0063] FIG. 4 shows an original image (400) for setting a region of interest according to embodiments of the present disclosure.

[0064] Referring to FIG. 4, a camera (110) is installed on a road near a vehicle driving direction change zone (210) and can capture an image (400) of the road near the vehicle driving direction change zone (210). An automatic region of interest setting device (120) according to embodiments of the present disclosure can receive image data for the image (400) captured by the camera (110) from the camera (110).

[0065] FIGS. 5 and 6 are drawings illustrating the process of generating a feature map for setting a region of interest according to embodiments of the present disclosure.

[0066] Referring to FIGS. 5 and 6, the characteristic map generating unit (320) of the automatic interest area setting device (120) according to embodiments of the present disclosure can generate vehicle movement path data for vehicles traveling on the road as a characteristic map based on image data of a road near a vehicle driving direction change area (210) captured through a camera (110).

[0067] Referring to FIGS. 5 and 6, the characteristic map generation unit (320) detects the time-series position (500) and movement direction (600) of each vehicle traveling on the road based on image data, and can generate vehicle movement path data including information on the time-series position (500) and movement direction (600) of each vehicle as a characteristic map. Accordingly, the characteristic map may include information on the time-series position and movement direction for each of the vehicles.

[0068] The characteristic map generation unit (320) can detect the direction of movement (600) of a vehicle by connecting the time-series positions (500) of the vehicle. The characteristic map generation unit (320) can generate a vehicle movement path as a characteristic map by accumulating the time-series positions (500) and the direction of movement (600) of each vehicle.

[0069] FIGS. 7 and 8 are drawings for illustrating a car recognition process for setting a region of interest according to embodiments of the present disclosure.

[0070] Referring to FIG. 7, the lane setting unit (330) of the automatic region of interest setting device (120) according to embodiments of the present disclosure can recognize N lanes (L1, L2, L3, L4) by performing continuity and clustering processing on the movement paths of each vehicle based on information regarding the time-dependent location and movement direction of each vehicle. Here, N may be a natural number greater than or equal to 1. In FIG. 7 and FIG. 8, the case where N is 4 is given as an example. That is, the example is a road with 4 lanes.

[0071] The lane setting unit (330) of the automatic interest area setting device (120) according to the embodiments of the present disclosure performs continuity and clustering processing for the movement paths of each of the vehicles, and based on the results of the continuity and clustering processing, can group the vehicles into a first driving direction vehicle group traveling in a first driving direction approaching the camera (110) and a second driving direction vehicle group traveling in a second driving direction moving away from the camera (110).

[0072] The lane setting unit (330) of the automatic interest area setting device (120) according to the embodiments of the present disclosure can subdivide a first driving direction vehicle group into N detailed vehicle groups based on the results of performing continuity and clustering processing, and set the movement path of each of the N detailed vehicle groups into N lanes (L1, L2, L3, L4).

[0073] The lane setting unit (330) of the automatic interest area setting device (120) according to the embodiments of the present disclosure may set the movement path of the detailed vehicle group closest to the second driving direction vehicle group among N detailed vehicle groups as the first lane (L1) among N lanes, and may set the (N-1) lanes (L2, L3, L4) excluding the first lane among N lanes in order of proximity to the first lane, from the second lane to the Nth lane (L2, L3, L4).

[0074] Referring to FIGS. 7 and 8, the lane setting unit (330) of the automatic interest area setting device (120) according to embodiments of the present disclosure may set an overlapping area (800) within each of the N lanes (L1, L2, L3, L4) and set a stop line (Ls) that is the boundary between the overlapping area (800) and the area (L1', L2', L3', L4') excluding the overlapping area (800) from each of the N lanes (L1, L2, L3, L4).

[0075] Referring to FIG. 8, the overlapping area (800) may correspond to an area where two or more different vehicle movement directions overlap. For example, the overlapping area (800) may be an intersection area.

[0076] FIG. 9 shows the result of setting the region of interest according to embodiments of the present disclosure.

[0077] Referring to FIG. 9, the interest area setting unit (340) of the interest area automatic setting device (120) according to embodiments of the present disclosure may distinguish the movement path and movement direction of a vehicle moving from a stop line (Ls) for each of N lanes (L1', L2', L3', L4'), and for each of N lanes (L1', L2', L3', L4'), include a previous interest area (ROI1) in the section before the vehicle reaches the stop line (Ls) and an arrival interest area (ROI2) that the vehicle will reach after passing the stop line (Ls).

[0078] A region of interest for each of the N lanes (L1', L2', L3', L4') may include a previous region of interest (ROI1) in the section prior to the vehicle driving direction change zone (210) being reached by the vehicle driving in the lane, and a destination region of interest (ROI2) being reached by the vehicle driving in the lane after passing through the vehicle driving direction change zone (210).

[0079] For example, the previous region of interest (ROI1) may include straight sections (910, 920, 930, 940) before the vehicle reaches the stop line (Ls) for each of the N lanes (L1', L2', L3', L4').

[0080] For example, the destination area of ​​interest (ROI2) may include at least one of a straight direction destination area of ​​interest (SA) that the vehicle will reach by driving in a straight direction in the vehicle driving direction change area (210), a first rotational direction area of ​​interest (LA) that the vehicle will reach by turning in a first rotational direction in the vehicle driving direction change area (210), a second rotational direction area of ​​interest (RA) that the vehicle will reach by turning in a second rotational direction different from the first rotational direction in the vehicle driving direction change area (210), and a U-turn direction area of ​​interest (UA) that the vehicle will reach by making a U-turn in the opposite direction in the vehicle driving direction change area (210).

[0081] In the foregoing, an automatic region of interest setting method performed by an automatic region of interest setting device (120) according to embodiments of the present disclosure has been described. Below, the automatic region of interest setting method according to the embodiments of the present disclosure described above will be briefly explained again.

[0082] FIG. 10 is a flowchart of a method for automatically setting a region of interest according to embodiments of the present disclosure.

[0083] Referring to FIG. 10, the method for automatically setting a region of interest according to embodiments of the present disclosure may include the step (S20) of generating vehicle movement path data for vehicles traveling on the road as a feature map of the road based on image data of a road near a vehicle driving direction change area (210) captured through a camera (110), the step (S30) of setting N lanes (N is a natural number greater than or equal to 1) included in the road based on the feature map, and the step (S40) of setting a region of interest for each of the N lanes based on the feature map.

[0084] Referring to FIG. 10, the method for automatically setting a region of interest according to embodiments of the present disclosure may further include a step (S10) of recognizing vehicles driving on a road in image data through deep learning prior to the step (S20) of generating a feature map.

[0085] In the step of generating a feature map (S20), vehicle movement path data for vehicles driving on the road can be obtained based on the vehicle recognition result of step S10.

[0086] Referring to FIG. 10, the method for automatically setting a region of interest according to embodiments of the present disclosure may further include, after the step of setting a region of interest (S40), a step (S50) of resetting a region of interest for each of N lanes when a region of interest reset event information is input.

[0087] For example, the interest area reset event information may include at least one of replacement information of the camera (110), information on changing the installation location of the camera (110), information on changing the viewing angle of the camera (110), and information on changing the settings of the camera (110).

[0088] The embodiments of the present disclosure described above are briefly explained as follows.

[0089] An automatic interest region setting device according to embodiments of the present disclosure may include a feature map generating unit that generates vehicle movement path data for vehicles traveling on the road as a feature map for the road based on image data of a road near a vehicle driving direction change zone captured through a camera, a lane setting unit that sets N lanes (N is a natural number greater than or equal to 1) included in the road based on the feature map, and an interest region setting unit that sets an interest region for each of the N lanes based on the feature map.

[0090] A region of interest for each of the N lanes may include a previous region of interest in the section prior to the vehicle traveling in the lane reaching the vehicle direction change zone, and a destination region of interest that the vehicle traveling in the lane will reach after passing through the vehicle direction change zone.

[0091] The arrival area of ​​interest may include at least one of a straight direction arrival area of ​​interest that the vehicle will reach by driving in a straight direction in the vehicle driving direction change area, a first rotational direction area of ​​interest that the vehicle will reach by turning in a first rotational direction in the vehicle driving direction change area, a second rotational direction area of ​​interest that the vehicle will reach by turning in a second rotational direction different from the first rotational direction in the vehicle driving direction change area, and a U-turn direction area of ​​interest that the vehicle will reach by making a U-turn in the opposite direction in the vehicle driving direction change area.

[0092] The characteristic map generation unit can generate vehicle movement path data as a characteristic map that includes information on the time-series location and movement direction for each of the vehicles.

[0093] The lane setting unit can recognize N lanes by performing continuity and clustering processing on the movement paths of each vehicle based on information regarding the time-dependent position and movement direction of each vehicle.

[0094] The lane setting unit performs continuity and clustering processing for the movement paths of each vehicle, and based on the results of the continuity and clustering processing, groups the vehicles into a first driving direction vehicle group traveling in a first driving direction approaching the camera and a second driving direction vehicle group traveling in a second driving direction moving away from the camera, subdivides the first driving direction vehicle group into N detailed vehicle groups, and sets the movement path of each of the N detailed vehicle groups into N lanes.

[0095] The lane setting unit can set the movement path of the detailed vehicle group closest to the second driving direction vehicle group among N detailed vehicle groups as the first lane among N lanes, and set the (N-1) lanes excluding the first lane among N lanes in order of proximity to the first lane, from the second lane to the Nth lane.

[0096] The lane setting unit sets an overlapping area within each of the N lanes, sets a stop line that is the boundary between the area excluding the overlapping area and the overlapping area from each of the N lanes, and the overlapping area may correspond to an area where two or more different vehicle movement directions overlap.

[0097] The interest region setting unit distinguishes the movement path and direction of a vehicle moving from a stop line for each of N lanes, and for each of N lanes, may include a previous interest region in the section before the vehicle reaches the stop line and an arrival interest region to be reached after the vehicle passes the stop line.

[0098] An automatic interest region setting device according to embodiments of the present disclosure may further include an interest region reset control unit that outputs a command signal for resetting the interest region for each of N lanes when interest region reset event information is input.

[0099] The interest area reset event information may include at least one of camera replacement information, camera installation location change information, camera field of view change information, and camera setting change information.

[0100] The vehicle driving direction change zone may be an intersection or a U-turn zone.

[0101] The region of interest automatic setting device according to the embodiments of the present disclosure may further include a deep learning-based vehicle recognition unit that recognizes a vehicle driving on a road in image data through deep learning.

[0102] The feature map generation unit can obtain vehicle movement path data for vehicles driving on the road based on the vehicle recognition results of the deep learning-based vehicle recognition unit.

[0103] A traffic volume measurement system according to embodiments of the present disclosure may include an automatic area of ​​interest setting device that sets an area of ​​interest for each of N lanes (N is a natural number greater than or equal to 1) included in the road based on image data of a road near a vehicle driving direction change zone captured through a camera, and a traffic volume measurement device that measures the traffic volume for each of the areas of interest for each of the N lanes.

[0104] The automatic interest region setting device may include a feature map generating unit that generates vehicle movement path data for vehicles traveling on the road as a feature map for the road based on image data, a lane setting unit that sets N lanes (N is a natural number greater than or equal to 1) included in the road based on the feature map, and an interest region setting unit that sets an interest region for each of the N lanes based on the feature map.

[0105] The camera may be installed on a road structure near the vehicle driving direction change zone. The automatic region of interest setting device may be installed on a road structure or on a road structure different from the road structure, or may be a traffic control center server that receives image data from the camera, or may be included in the traffic control center server.

[0106] A region of interest for each of the N lanes may include a previous region of interest in the section prior to the vehicle traveling in the lane reaching the vehicle direction change zone, and a destination region of interest that the vehicle traveling in the lane will reach after passing through the vehicle direction change zone.

[0107] The region of interest automatic setting device may further include a deep learning-based vehicle recognition unit that recognizes a vehicle driving on the road in image data through deep learning.

[0108] The feature map generation unit can obtain vehicle movement path data for vehicles driving on the road based on the vehicle recognition results of the deep learning-based vehicle recognition unit.

[0109] A method for automatically setting a region of interest according to embodiments of the present disclosure may include the steps of: acquiring vehicle movement path data for vehicles traveling on a road based on image data of a road near a vehicle driving direction change zone captured through a camera, and generating a feature map for the road based on the vehicle movement path data; setting N lanes (N is a natural number greater than or equal to 1) included in the road based on the feature map; and setting a region of interest for each of the N lanes based on the feature map.

[0110] The method for automatically setting a region of interest according to embodiments of the present disclosure may further include, prior to the step of generating a feature map, a step of recognizing vehicles driving on a road in image data through deep learning.

[0111] In the step of generating a feature map, the region of interest automatic setting device can acquire vehicle movement path data for vehicles traveling on the road based on vehicle recognition results.

[0112] The method for automatically setting a region of interest according to embodiments of the present disclosure may further include, after the step of setting a region of interest, a step of resetting a region of interest for each of N lanes when region of interest reset event information is input.

[0113] According to embodiments of the present disclosure, a method for automatically setting an area of ​​interest and an apparatus for automatically setting an area of ​​interest can be provided, which can accurately and quickly automatically set an area of ​​interest for measuring traffic volume.

[0114] According to embodiments of the present disclosure, a method for automatically setting an area of ​​interest and an apparatus for automatically setting an area of ​​interest can be provided, which can automatically reset the area of ​​interest when the angle of view of a camera changes.

[0115] According to embodiments of the present disclosure, a traffic volume measurement system can be provided that can accurately and quickly measure traffic volume by accurately and quickly setting an area of ​​interest.

[0116] The foregoing description is merely an illustrative explanation of the technical concept of the present disclosure, and those skilled in the art to which the present disclosure pertains may make various modifications and variations within the scope of the essential characteristics of the present disclosure. Furthermore, since the embodiments disclosed in the present disclosure are intended to explain, not limit, the scope of the technical concept of the present disclosure is not limited by these embodiments.

[0117]

[0118] [Explanation of the symbol]

[0119] 100: Traffic volume measurement system

[0120] 110: Camera

[0121] 120: Automatic region of interest setting device

[0122] 130: Traffic volume measuring device

[0123]

[0124] CROSS-REFERENCE TO RELATED APPLICATION

[0125] This patent application claims priority pursuant to Section 119(a) of the U.S. Patent Act (35 USC §119(a)) to Korean Patent Application No. 10-2024-0128611 filed on September 24, 2024, the entire contents of which are incorporated by reference into this patent application. Additionally, this patent application claims priority in countries other than the United States for the same reasons as above, and the entire contents of which are incorporated by reference into this patent application.

Claims

1. A feature map generating unit that generates vehicle movement path data for vehicles traveling on the road as a feature map for the road, based on image data of the road near the vehicle driving direction change zone captured through a camera; A lane setting unit that sets N lanes (N is a natural number greater than or equal to 1) included in the road based on the above characteristic map; and An automatic interest region setting device comprising an interest region setting unit that sets an interest region for each of the N lanes based on the above characteristic map.

2. In Paragraph 1, The above characteristic map generation unit is, An automatic region of interest setting device that generates the vehicle movement path data, including information on the time-series position and direction of movement for each of the above vehicles, as the characteristic map.

3. In Paragraph 2, The above lane setting unit is, An automatic interest region setting device that recognizes N lanes by performing continuity and clustering processing on the movement paths of each of the vehicles based on information regarding the time-dependent position and movement direction of each of the vehicles.

4. In Paragraph 3, The above lane setting unit is, Continuity and clustering processing is performed on the movement paths of each of the above vehicles, and Based on the results of the above continuity and clustering processing, The vehicles are grouped into a first driving direction vehicle group traveling in a first driving direction approaching the camera and a second driving direction vehicle group traveling in a second driving direction moving away from the camera. An automatic interest area setting device that subdivides the first driving direction vehicle group into N detailed vehicle groups and sets the movement path of each of the N detailed vehicle groups into the N lanes.

5. In Paragraph 4, The above lane setting unit is, Among the above N detailed vehicle groups, the movement path of the detailed vehicle group closest to the above second driving direction vehicle group is set as the first lane among the above N lanes, and An automatic interest area setting device that sets (N-1) lanes among the N lanes excluding the first lane, in order of proximity to the first lane, and sets the second lane to the Nth lane in sequence.

6. In Paragraph 1, An automatic interest area setting device further comprising an interest area reset control unit that outputs a command signal for resetting the interest area for each of the N lanes when interest area reset event information is input.

7. In Paragraph 6, The above interest area reset event information is, Replacement information for the above camera; Information on changing the installation location of the above camera; Information on changing the field of view of the above camera; and An automatic region of interest setting device comprising at least one of the setting change information of the above-mentioned camera.

8. In Paragraph 1, An automatic area of ​​interest setting device in which the above vehicle driving direction change area is an intersection or a U-turn area.

9. In Paragraph 1, It further includes a deep learning-based vehicle recognition unit that recognizes a vehicle driving on the road in the image data through deep learning, and The above characteristic map generation unit is, An automatic region of interest setting device that acquires vehicle movement path data for vehicles traveling on the road based on the vehicle recognition results of the deep learning-based vehicle recognition unit.

10. An automatic interest region setting device that sets an interest region for each of N (N is a natural number greater than or equal to 1) lanes included in the road based on image data of the road near the vehicle driving direction change zone captured through a camera; and It includes a traffic volume measuring device that measures the traffic volume for a region of interest for each of the N lanes mentioned above, The above-mentioned automatic interest area setting device is, A feature map generation unit that generates vehicle movement path data for vehicles traveling on the road based on the above image data as a feature map for the road; A lane setting unit that sets N lanes (N is a natural number greater than or equal to 1) included in the road based on the above characteristic map; and A traffic volume measurement system comprising an interest region setting unit that sets an interest region for each of the N lanes based on the above characteristic map.

11. In Paragraph 10, The above camera is installed on a road structure near the vehicle driving direction change area, and The above-mentioned automatic interest area setting device is, Installed on the above road structure or on a road structure different from the above road structure, A traffic control center server that receives the image data from the camera, or a traffic volume measurement system included in the traffic control center server.

12. In Paragraph 10, The above-described region of interest automatic setting device further includes a deep learning-based vehicle recognition unit that recognizes a vehicle driving on the road in the image data through deep learning, and The above characteristic map generation unit is, A traffic volume measurement system that acquires vehicle movement path data for vehicles traveling on the road based on the vehicle recognition results of the deep learning-based vehicle recognition unit.

13. A step of acquiring vehicle movement path data for vehicles traveling on the road based on image data of the road near the vehicle driving direction change zone captured through a camera, and generating a feature map for the road based on the vehicle movement path data; Based on the above characteristic map, the step of setting N lanes (N is a natural number greater than or equal to 1) included in the road; and A method for automatically setting a region of interest, comprising the step of setting a region of interest for each of the N lanes based on the above characteristic map.

14. In Paragraph 13, Prior to the step of generating the above feature map, the method further includes the step of recognizing vehicles driving on the road in the image data through deep learning, and A method for automatically setting a region of interest, wherein in the step of generating the above feature map, vehicle movement path data for vehicles traveling on the road is obtained based on vehicle recognition results.

15. In Paragraph 13, After the step of setting the region of interest mentioned above, A method for automatically setting an area of ​​interest, further comprising the step of resetting the area of ​​interest for each of the N lanes when information on an area of ​​interest reset event is input.

Images