JP7195200B2 - In-vehicle device, in-vehicle system, and surrounding monitoring method - Google Patents

Description

The disclosed embodiments relate to an in-vehicle device, an in-vehicle system, and a surroundings monitoring method.

2. Description of the Related Art Conventionally, there has been known a technique of detecting an object existing in the moving direction of a vehicle at an intersection or the like with a peripheral monitoring sensor such as a camera (see, for example, Patent Document 1).

JP 2015-230567 A

However, the conventional technology described above has room for further improvement in terms of improving the object detection accuracy.

For example, when making a right or left turn at an intersection, the crosswalk that crosses where the vehicle is turning is different from the direction of movement of the vehicle before the vehicle begins to turn, during the turn, and after the vehicle finishes turning. is different.

For this reason, for example, when trying to detect an object based on an image from the front camera, if a fixed area on the image that always includes a pedestrian crossing that looks different as described above is set as a detection target area, There is a risk of increased frequency of erroneous detection, in which objects that do not need to be detected, and undetected objects, in which objects are overlooked.

One aspect of the embodiments has been made in view of the above, and an object thereof is to provide an in-vehicle device, an in-vehicle system, and a surroundings monitoring method capable of improving object detection accuracy.

An in-vehicle device according to an aspect of an embodiment is an in-vehicle device that detects objects around a vehicle based on images captured by a plurality of cameras mounted on the vehicle, and includes an acquisition unit, a determination unit, and an angle calculator. a detection control unit; and a detection unit. The acquisition unit acquires a running condition of the vehicle. The determination unit determines an object detection target area according to the traveling situation acquired by the acquisition unit. The angle calculation unit calculates an entry angle of the vehicle with respect to the detection target area according to the detection target area and the running condition determined by the determination unit. The detection control unit changes a detection condition regarding object detection according to the approach angle calculated by the angle calculation unit. The detection unit detects an object based on the detection condition changed by the detection control unit and the image captured by the camera.

According to one aspect of the embodiment, it is possible to improve the object detection accuracy.

FIG. 1A is a schematic explanatory diagram (part 1) of a surroundings monitoring method according to an embodiment; FIG. 1B is a schematic explanatory diagram (part 2) of the surroundings monitoring method according to the embodiment; FIG. 1C is a schematic explanatory diagram (part 3) of the surroundings monitoring method according to the embodiment; FIG. 1D is a schematic explanatory diagram (part 4) of the surroundings monitoring method according to the embodiment; FIG. 2 is a block diagram of an in-vehicle system according to the embodiment. FIG. 3A is a diagram (part 1) illustrating a specific example of changing a detection condition according to the embodiment; FIG. 3B is a diagram (part 2) illustrating a specific example of changing the detection condition according to the embodiment; FIG. 3C is a diagram (part 3) illustrating a specific example of changing the detection condition according to the embodiment; FIG. 3D is a diagram (part 4) illustrating a specific example of changing the detection condition according to the embodiment; FIG. 4A is a diagram (part 1) showing a specific example of changing the detection condition according to the modification; FIG. 4B is a diagram (part 2) illustrating a specific example of changing the detection condition according to the modification; FIG. 4C is a diagram (part 3) showing a specific example of changing the detection condition according to the modification; FIG. 5 is a flowchart illustrating a processing procedure executed by the in-vehicle device according to the embodiment;

Hereinafter, embodiments of an in-vehicle device, an in-vehicle system, and a surroundings monitoring method disclosed in the present application will be described in detail with reference to the accompanying drawings. In addition, this invention is not limited by embodiment shown below.

1A to 1D, and then the in-vehicle system 1 to which the surroundings monitoring method according to the embodiment is applied will be described with reference to FIGS. 2 to 5. I will explain.

First, an overview of the surroundings monitoring method according to the embodiment will be described with reference to FIGS. 1A to 1D. 1A to 1D are schematic explanatory diagrams (part 1) to (part 4) of the surroundings monitoring method according to the embodiment.

In addition, below, as shown to FIG. 1A, it demonstrates assuming the case where the vehicle V turns right in an intersection. As shown in FIG. 1A, in a scene in which vehicle V turns right, for example, bicycle B may approach a pedestrian crossing ahead of vehicle V turning. The surroundings monitoring method according to the embodiment is intended to improve detection accuracy in such a case, for example, a pedestrian crossing, as a detection target region DR in which an object is to be mainly detected.

However, as shown in FIG. 1B, in each scene in which the vehicle V turns right, if the object detection range is set to include, for example, the entire image or the fixed region FR so as to include all the detection target regions DR that change from moment to moment, There is a risk of increasing the chances of false detections and non-detections.

For example, in the upper scene of FIG. 1B, if a sidewalk opposite to the direction in which the vehicle V turns is included in the fixed area FR as indicated by M1 in the figure, a pedestrian or the like walking on the sidewalk is detected. There is a risk of doing so.

Further, for example, in the scene in the middle of FIG. 1B, if a pedestrian crossing that is not included in the detection target region DR is included in the fixed region FR as indicated by M2 in the drawing, There is a risk of detecting a person, etc.

Further, for example, in the scene in the lower part of FIG. 1C, if the fixed region FR includes a footpath that is not included in the detection target region DR beyond which the vehicle V turns, as indicated by M3 in the drawing, such There is a risk of detecting a pedestrian or the like walking on the sidewalk.

Therefore, in the surroundings monitoring method according to the embodiment, the detection target area DR is determined according to the running condition of the vehicle V, the approach angle of the vehicle V to the detection target area DR is calculated, and according to the calculated approach angle, We decided to change the object detection conditions.

Note that the detection conditions referred to here are, for example, the weighting of each of the plurality of peripheral monitoring cameras mounted on the vehicle V, the detection strength (can also be referred to as "reliability"), and the like.

Specifically, as shown in FIG. 1C, in the surroundings monitoring method according to the embodiment, a plurality of cameras 4 mounted on a vehicle V are used. The cameras 4 include, for example, a front camera 4-F whose imaging range is the front of the vehicle V, a right side camera 4-R whose imaging range is the right side of the vehicle V, and a left camera 4-R whose imaging range is the left side of the vehicle V. A side camera 4-L and a back camera 4-B whose imaging range is the rear of the vehicle V. FIG. It should be noted that FIG. 1C does not limit the mounting location or imaging range of each camera 4 .

Then, in the surroundings monitoring method according to the embodiment, first, the running condition of the vehicle V is acquired as needed. Such driving conditions include, for example, a predicted route based on navigation information, current position, vehicle speed, manner of blinker operation, steering angle, and the like. Then, in the surroundings monitoring method according to the embodiment, the detection target area DR is determined based on the acquired driving conditions.

Then, in the surroundings monitoring method according to the embodiment, the approach angle of the vehicle V to the determined detection target area DR is calculated. Here, as shown in FIG. 1D, the "approach angle" is an angle θ formed between the determined "extending direction of the detection target region DR" and the "advancing direction of the vehicle V". Therefore, the approach angle θ gradually increases from an acute angle until the vehicle V finishes turning. Then, in the surroundings monitoring method according to the embodiment, the object detection conditions are dynamically changed according to the change in the approach angle θ.

Here, the extending direction of the detection target region DR can be considered as a direction along the shape of the detection target region DR in terms of physical shape. For example, if the detection target area DR is a pedestrian crossing, a straight line connecting the sidewalk at the entrance and the sidewalk at the exit can be set as the extending direction. Further, if the detection target region DR is a road, the direction along the road can be set as the extending direction.

Further, the extending direction of the detection target region DR can also be considered to be the direction in which the detection target mainly moves in the detection target region DR from a functional semantic point of view. For example, if the detection target area DR is a pedestrian crossing, the extension direction can be the direction in which the pedestrian or bicycle, which is the detection target, mainly moves. Further, if the detection target area DR is a road, the direction in which the vehicle, which is the detection target, mainly moves can be set as the extending direction.

In other words, the approach angle can be considered to be the angle at which the travel direction of the own vehicle intersects with the movement direction of the assumed object to be detected.

For example, in the surroundings monitoring method according to the embodiment, as an example of changing such detection conditions, as shown in FIG. , and the front camera 4-F is weighted as a secondary “sub” camera. In other words, the right side camera 4-R is selected as the main camera, the front camera 4-F is selected as the sub camera, and the other cameras 4 are not selected.

In addition, in the surroundings monitoring method according to the embodiment, an object is detected while using the imaged images of each area covered by the imaging range of each of the right side camera 4-R and the front camera 4-F in accordance with the weighting described above. .

As a result, it is possible to detect the detection target area DR according to the change in the approach angle θ using the optimum image captured by the camera 4, and it is possible to detect the object with high accuracy while following the blind spot of the driver of the vehicle V. . A more specific example of changing the detection conditions will be described later with reference to FIGS. 3A to 4C.

As described above, in the surroundings monitoring method according to the embodiment, the detection target region DR is determined according to the running condition of the vehicle V, the approach angle θ of the vehicle V with respect to the detection target region DR is calculated, and the approach angle θ is calculated. The object detection conditions, including the weighting of each camera 4, are dynamically changed in accordance with changes in .

Therefore, according to the surroundings monitoring method according to the embodiment, it is possible to improve the detection accuracy of an object existing in the direction in which the vehicle V is moving. Hereinafter, a configuration example of the vehicle-mounted system 1 including the vehicle-mounted device 10 to which the surroundings monitoring method according to the embodiment is applied will be described more specifically with reference to FIG. 2 and subsequent figures.

FIG. 2 is a block diagram of the in-vehicle system 1 according to the embodiment. In addition, in FIG. 2, only the components necessary for explaining the features of the present embodiment are represented by functional blocks, and the description of general components is omitted.

In other words, each component illustrated in FIG. 2 is functionally conceptual and does not necessarily need to be physically configured as illustrated. For example, the specific forms of distribution and integration of each functional block are not limited to those shown in the figure, and all or part of them can be functionally or physically distributed in arbitrary units according to various loads and usage conditions.・It is possible to integrate and configure.

As shown in FIG. 2, an in-vehicle system 1 includes various sensors including a GPS (Global Positioning System) sensor 2 and a steering angle sensor 3, cameras 4 such as a front camera 4-F, a right side camera 4-R, a left side A camera 4-L, a back camera 4-B, an in-vehicle device 10, and a vehicle control device 20 are included.

The in-vehicle device 10 includes a storage section 11 and a control section 12 . The in-vehicle device 10 communicates with a vehicle control device 20 that controls various behaviors of the vehicle V, the GPS sensor 2, the steering angle sensor 3, the camera 4, and the like via a communication network such as a CAN (Controller Area Network). communicatively connected.

The vehicle control device 20 performs vehicle control such as PCS (Pre-crash Safety System) and AEB (Advanced Emergency Braking System) based on the object detection result by the in-vehicle device 10 . Further, the vehicle control device 20 may be an automatic driving control device that performs automatic driving control based on the detection result of the object by the in-vehicle device 10 .

The storage unit 11 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disc, and stores navigation information 11a in the example of FIG. do. The navigation information 11a is information relating to the car navigation system of the vehicle V, and includes map information, a planned travel route, and the like.

The control unit 12 is a controller, and various programs stored in a storage device inside the in-vehicle device 10 are executed by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like, using the RAM as a work area. It is realized by being executed. Also, the control unit 12 can be implemented by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

The control unit 12 includes an acquisition unit 12a, a determination unit 12b, an angle calculation unit 12c, a detection control unit 12d, and a detection unit 12e, and implements or executes information processing functions and actions described below. .

The acquisition unit 12a acquires the traveling condition of the vehicle V based on the sensor values of various sensors such as the GPS sensor 2 and the steering angle sensor 3, and the navigation information 11a. The traveling condition of the vehicle V includes, for example, the position of the vehicle V, the direction of travel, and the route setting conditions such as right and left turns. In addition, the acquisition unit 12a notifies the determination unit 12b of the acquired driving conditions.

The determination unit 12b determines the above-described detection target region DR based on the running condition of the vehicle V acquired by the acquisition unit 12a. For example, as shown in FIG. 1A, when the vehicle V is entering the intersection and the route is set to turn right, the pedestrian crossing on the right side of the intersection as viewed from the vehicle V is determined as the detection target area DR. Note that once the detection target region DR is determined, it is preferable not to change it until a predetermined condition is satisfied. The predetermined condition is, for example, that a predetermined time has passed since the determination, that the vehicle travels a predetermined distance after the determination, that the vehicle crosses the detection target region DR, or the like. Further, the determination unit 12b notifies the detected detection target region DR to the angle calculation unit 12c.

The angle calculator 12c calculates the above-described approach angle θ based on the driving situation acquired by the acquirer 12a and the detection target region DR determined by the determiner 12b. Further, the angle calculator 12c updates the approach angle θ each time based on changes in the traveling conditions acquired by the acquirer 12a. Further, the angle calculation unit 12c notifies the detection control unit 12d of the calculated approach angle θ.

The detection control unit 12d dynamically changes the detection conditions according to the approach angle θ calculated by the angle calculation unit 12c.

The detection unit 12e detects an object based on the captured image from the camera 4 while complying with the detection conditions changed by the detection control unit 12d. Further, the detection unit 12e notifies the vehicle control device 20 of the detection result, for example.

Note that the detection unit 12e may output the detection result to an output unit of the in-vehicle device 10 (not shown). The output unit is, for example, a display or a speaker provided inside the vehicle.

A specific example of changing the detection condition according to the embodiment will now be described with reference to FIGS. 3A to 3D. 3A to 3D are diagrams (part 1) to (part 4) showing specific examples of changing detection conditions according to the embodiment.

Consider a case where the approach angle θ calculated by the angle calculator 12c is smaller than a predetermined angle, as shown in FIG. 3A. For example, if the predetermined angle is 50° and the approach angle θ is “0°≦θ<20°”, the detection control unit 12d weights the right side camera 4-R as “main” and the front camera 4 - Weight F as "sub".

That is, when the approach angle θ is small and the detection target area DR is on the side of the inner wheel of the vehicle V, the right side camera 4-R whose main imaging range is on the side of the inner wheel of the vehicle V is defined as the "main" camera. weight. As a result, the detection target region DR can be contained within the main imaging range of the camera 4 weighted as "main". On the other hand, the front camera 4-F, which may capture the detection target area DR but is highly likely not to be within the main imaging range, is weighted as "sub". Note that the main imaging range refers to a range in which detection processing can be accurately performed in the image captured by the camera 4 . This is, for example, near the center of the image.

Then, the detection control unit 12d causes the detection unit 12e to use the captured image of each area covered by the imaging range of each of the right side camera 4-R and the front camera 4-F in accordance with the weighting, and causes the detection unit 12e to detect the object. .

Specifically, when the detection unit 12e detects an object, it is preferable to change the reliability when integrating the detection results of the images captured by the plurality of cameras 4. FIG. For example, the reliability of the detection result based on the image captured by the camera 4 weighted as "main" is set higher than the reliability of the detection result based on the image captured by the camera 4 weighted as "sub".

In addition, it is preferable to change the order in which detection processing is performed for each of the images captured by the plurality of cameras 4 . For example, an image captured by the camera 4 weighted as "main" is subjected to detection processing before an image captured by the camera 4 weighted as "sub".

In addition, it is preferable to change the frequency of detection processing for each of the images captured by the plurality of cameras 4 . For example, an image captured by the camera 4 weighted as "main" is subjected to detection processing more frequently than an image captured by the camera 4 weighted as "sub". Specifically, in the case of "main" weighting, detection is performed every frame, and in the case of "sub" weighting, detection is performed every several frames.

Also, for example, zero can be set as the weight to be set. The image captured by the camera 4 set to zero may be processed so as not to be used for object detection in the detection unit 12e. In this case, in other words, it can be said that the camera 4 to be used for object detection and the camera 4 not to be used are set in the detection unit 12e. For example, in the case of FIG. 3A, there is a high possibility that the detection target area DR will not be captured by the left side camera 4-L or the rear camera 4-B. Therefore, there is no problem even if the detector 12e does not use it for detection. Also, the camera 4 set as "sub" like the front camera 4-F may not be used for detection because there is a high possibility that the detection target area DR cannot be included in the main imaging range.

Also, as shown in FIG. 3B, consider a case where the approach angle θ calculated by the angle calculator 12c has increased but is still smaller than a predetermined angle, for example, "20°≦θ<50°". Also in this case, the detection control unit 12d weights the right side camera 4-R as "main" and weights the front camera 4-F as "sub".

In the case of FIG. 3B, each area covered by the imaging range of each of the right side camera 4-R and the front camera 4-F rotates clockwise as compared to the case of FIG. 3A. Then, similarly to the case of FIG. 3A, the detection control unit 12d causes the detection unit 12e to detect the object by using the captured image of each region according to the above-described weighting.

Also, as shown in FIG. 3C, consider a case where the approach angle θ calculated by the angle calculator 12c exceeds a predetermined angle and becomes close to a right angle. For example, when the approach angle θ is “50°≦θ<80°”, the detection control unit 12d weights the front camera 4-F as “main” and the right side camera 4-R as “sub”. and weighted. That is, when the approach angle θ approaches a right angle and the detection target area DR comes in front of the vehicle V, the front camera 4-F whose main imaging range is in front of the vehicle V is weighted as "main". As a result, the detection target region DR can be contained within the main imaging range of the camera 4 weighted as "main". On the other hand, the right side camera 4-R, which may capture the detection target area DR but is highly likely not to be within the main imaging range, is weighted as "sub".

Of course, in the case of FIG. 3C, each region covered by the imaging range of each of the right side camera 4-R and the front camera 4-F rotates further clockwise than in the case of FIG. 3B. 3A and 3B, the detection control unit 12d causes the detection unit 12e to detect the object by using the captured image of each area according to the above-described weighting.

Further, as shown in FIG. 3D, when the approach angle θ calculated by the angle calculation unit 12c is close to a right angle, for example, “80°≦θ<120°”, the detection control unit 12d detects the front camera 4-F. is weighted as "main", and the right side camera 4-R and left side camera 4-L are weighted as "sub".

Needless to say, in the case of FIG. 3D, each area covered by the imaging range of the right side camera 4-R and the front camera 4-F rotates further clockwise than in the case of FIG. 3C. Also, an area covered by the imaging range of the left side camera 4-L is added. 3A to 3C, the detection control unit 12d causes the detection unit 12e to detect the object by using the captured image of each area according to the above-described weighting.

Here, the predetermined angle may be determined based on which camera 4 has a main imaging range in which the detection target region DR is mainly captured. For example, in the above description, the predetermined angle is 50° as an example. It may be considered that the detection target region DR is mainly captured in the main imaging range of the front camera 4-F rather than the main imaging range of the right side camera 4-R at the boundary of 50°. The predetermined angle should be set in advance based on the optical system of the camera 4 or the like. Alternatively, it may be changed depending on the distance between the detection target area DR and the vehicle V. FIG.

It should be noted that specific examples of changing detection conditions according to the embodiment are not limited to those shown in FIGS. 3A to 3D. Next, a specific example of changing the detection condition according to the modified example will be described with reference to FIGS. 4A to 4C. 4A to 4C are diagrams (Part 1) to (Part 3) showing specific examples of changes in detection conditions according to the modification.

By the way, when detecting an object, the detection unit 12e extracts an area (hereinafter referred to as "applicable area AR") corresponding to the detection target area DR determined by the determination unit 12b in the image captured by the camera 4, and extracts such area. Feature points of the object such as the position and velocity vector of the object are extracted in the relevant area AR. However, in extracting such feature points, using a uniform constant for the entire relevant area AR can be a factor in lowering the processing efficiency.

Therefore, as shown in FIG. 4A, for example, in the relevant area AR, for example, an area AR-1 corresponding to the edge of the area, that is, the area in which the object is newly captured in the imaging range of the camera 4, is focused on the feature points. can be extracted, the processing efficiency can be improved. Specifically, as shown in FIG. 4A, the detection control unit 12d increases the weighting of the area AR-1 into which the object such as the bicycle B enters, and the detection unit 12e for the area AR-1. Feature points can be extracted with priority.

Also, a moving object is easy to detect if, for example, there is a lot of horizontal vector movement. Therefore, as shown in FIG. 4B, the detection control unit 12d may select the camera 4 that captures an image with many horizontal vector movements as the "main" according to the approach angle θ. This makes it easier to detect moving objects.

Further, as shown in FIG. 4C, the detection control unit 12d may change the detection intensity of each camera 4 in accordance with the change in the approach angle θ from 0° to 120° when turning right, for example. . In the above description, the two cameras 4 are weighted in two stages of "main" and "sub" as the detection strength. On the other hand, for example, as shown in FIG. 4C, the detection control unit 12d sets the detection strength of the front camera 4-F to "weak" in response to the change in the approach angle θ from 0° to 120° when turning right. ” → “Medium” → “Strong”.

Further, the detection control unit 12d may change the detection strength of the right side camera 4-R from "medium"→"strong"→"strong". Further, the detection control unit 12d may change the detection strength of the left side camera 4-L from "weak"→"medium"→"strong".

Note that unlike the previous example, even when the approach angle θ is close to a right angle, even if the detection strength of the right side camera 4-R and the left side camera 4-L, which are side cameras, is set high, good. This is because, for example, when the detection target area DR is a pedestrian crossing, the front camera 4-F detects front and side objects such as entrances and exits of the pedestrian crossing, which are at the ends of the imaging range.

It should be noted that the detection strength may be rephrased as "reliability" as described above. It can also be called "priority". Such detection strength can be used, for example, as a coefficient when comprehensively judging the detection results based on the images captured by the respective cameras 4, thereby contributing to accurate object detection according to the approach angle θ. .

Next, a processing procedure executed by the in-vehicle device 10 according to the embodiment will be described with reference to FIG. 5 . FIG. 5 is a flow chart showing a processing procedure executed by the in-vehicle device 10 according to the embodiment.

As shown in FIG. 5, first, the acquisition unit 12a acquires the driving situation (step S101). Then, the determination unit 12b determines the detection target region DR according to the acquired driving conditions (step S102).

Then, the angle calculation unit 12c calculates an approach angle θ with respect to the detection target area DR based on the determined detection target area DR and the traveling situation (step S103). Then, the detection control unit 12d changes the detection conditions according to the calculated approach angle θ (step S104).

Then, the detection unit 12e detects an object according to the detection condition changed by the detection control unit 12d (step S105), and repeats the processing from step S101.

As described above, the in-vehicle device 10 according to the embodiment is the in-vehicle device 10 that detects objects around the vehicle V based on images captured by the plurality of cameras 4 mounted on the vehicle V, and the acquisition unit 12a , a determination unit 12b, an angle calculation unit 12c, a detection control unit 12d, and a detection unit 12e. Acquisition unit 12a acquires the running condition of vehicle V. FIG. The determination unit 12b determines the object detection target region DR according to the traveling situation acquired by the acquisition unit 12a. The angle calculation unit 12c calculates an entry angle θ of the vehicle V with respect to the detection target area DR according to the detection target area DR and the running condition determined by the determination unit 12b. The detection control unit 12d changes the detection conditions for object detection according to the approach angle θ calculated by the angle calculation unit 12c. The detection unit 12 e detects an object based on the detection conditions changed by the detection control unit 12 d and the image captured by the camera 4 .

Therefore, according to the in-vehicle device 10 according to the embodiment, it is possible to improve the object detection accuracy.

Further, the angle calculator 12c calculates an angle formed between the extending direction of the detection target region DR and the traveling direction of the vehicle V as an approach angle θ.

Therefore, according to the in-vehicle device 10 according to the embodiment, it is possible to accurately detect an object existing in the moving direction of the vehicle V when the vehicle V turns right or left.

Further, the detection control unit 12d weights each image captured by the camera 4 based on the imaging range and the approach angle θ of each camera 4 when the detection unit 12e detects an object.

Therefore, according to the in-vehicle device 10 according to the embodiment, the camera 4 that can capture the detection target region DR within the main imaging range may be weighted as "main", or the detection target region DR may be captured. It is possible to set detection conditions suitable for imaging the detection target area DR, such as weighting the cameras 4 that are highly likely not to contain the detection target area DR within the main imaging range as "sub".

The detection control unit 12d also compares the images captured by the cameras 4, and selects the camera 4 that captured images with relatively large horizontal vector movements.

Therefore, according to the in-vehicle device 10 according to the embodiment, it is possible to improve the detection accuracy of a moving object.

Further, the detection control unit 12d causes the detection unit 12e to selectively extract the feature points of the object with respect to the area AR-1 corresponding to the edge of the area corresponding to the detection target area DR in the image captured by the camera 4. .

Therefore, according to the in-vehicle device 10 according to the embodiment, it is possible to detect a moving object with high processing efficiency.

Further, the detection control unit 12d increases the detection intensity of the camera 4 whose main imaging range is the front of the vehicle V as the approach angle θ approaches a right angle.

Therefore, according to the in-vehicle device 10 according to the embodiment, as the movement direction of the vehicle V becomes perpendicular to the detection target area DR, the detection strength of the front camera 4-F, which is optimal for imaging the detection target area DR, is increased. Thereby, the object detection accuracy can be improved.

Further, when the approach angle θ is smaller than a predetermined angle, the detection control unit 12d captures the side of the vehicle V, which is closer to the inner wheels than the detection intensity of the camera 4 whose main imaging range is the front of the vehicle V. The detection strength of the camera 4 to be covered is increased.

Therefore, according to the in-vehicle device 10 according to the embodiment, when the detection target region DR is located on the side of the vehicle V, the right side camera 4-R or the left side camera 4-R is optimal for imaging the detection target region DR. Object detection accuracy can be improved by increasing the detection strength of the side camera 4-L.

In the above-described embodiment, right-turning is taken as an example, but left-turning is only bilaterally symmetrical with right-turning, and the above-described embodiment can be similarly applied to left-turning.

In addition, in the above-described embodiment, an intersection was taken as an example, but of course, the above-described embodiment can be applied in various scenes such as when entering a convenience store from a public road, when moving in a parking lot, and the like. can be done.

In addition, in the above-described embodiment, four cameras 4 are provided, but the number may be two or three, or may be five or more.

Further effects and modifications can be easily derived by those skilled in the art. Therefore, the broader aspects of the invention are not limited to the specific details and representative embodiments so shown and described. Accordingly, various changes may be made without departing from the spirit or scope of the general inventive concept defined by the appended claims and equivalents thereof.

1 in-vehicle system 4 camera 4-B back camera 4-F front camera 4-L left side camera 4-R right side camera 10 in-vehicle device 11a navigation information 12a acquisition unit 12b determination unit 12c angle calculation unit 12d detection control unit 12e detection unit 20 Vehicle control device DR Detection target area V Vehicle θ Approach angle

Claims (9)

An in-vehicle device that detects objects around the vehicle based on images captured by a plurality of cameras mounted on the vehicle,
an acquisition unit that acquires the running status of the vehicle;
a determination unit that determines an object detection target area according to the traveling situation acquired by the acquisition unit;
an angle calculation unit that calculates an entry angle of the vehicle with respect to the detection target area according to the detection target area and the running situation determined by the determination unit;
a detection control unit that changes detection conditions related to object detection according to the approach angle calculated by the angle calculation unit;
and a detection unit that detects an object based on the detection condition changed by the detection control unit and an image captured by the camera. The angle calculator,
The in-vehicle device according to claim 1, wherein an angle formed by an extending direction of the detection target area and a traveling direction of the vehicle is calculated as the approach angle. The detection control unit is
3. The detecting unit according to claim 1 or 2, wherein weighting is applied to each of the images captured by the cameras based on the imaging range and the approach angle of each of the cameras when the detection unit detects an object. In-vehicle device. The detection control unit is
4. The in-vehicle device according to claim 1, wherein the images captured by the respective cameras are compared, and the camera that captures images with relatively large lateral vector movements is selected. The detection control unit is
5. The method according to any one of claims 1 to 4, wherein the detection unit is caused to extract feature points of an object intensively with respect to an area corresponding to an end of the area corresponding to the detection target area in the image captured by the camera. The in-vehicle device according to any one of the above. The detection control unit is
The in-vehicle device according to any one of claims 1 to 5, wherein as the approach angle approaches a right angle, the detection strength of the camera whose main imaging range is the front of the vehicle is increased. The detection control unit is
When the approach angle is smaller than a predetermined angle, the detection strength of the camera whose imaging range is the side of the vehicle, which is the inner wheel side, is higher than the detection strength of the camera whose main imaging range is the front of the vehicle. The in-vehicle device according to any one of claims 1 to 6, characterized in that: an in-vehicle device according to any one of claims 1 to 7;
An in-vehicle system comprising: an in-vehicle sensor capable of outputting sensor information indicating the driving situation, the in-vehicle system including a plurality of the cameras. A surroundings monitoring method using an in-vehicle device for detecting objects around the vehicle based on images captured by a plurality of cameras mounted on the vehicle,
an acquisition step of acquiring the running status of the vehicle;
a determination step of determining an object detection target area according to the traveling situation acquired by the acquisition step;
an angle calculation step of calculating an entry angle of the vehicle with respect to the detection target region according to the detection target region and the running situation determined by the determination step;
a detection control step of changing a detection condition related to object detection according to the approach angle calculated by the angle calculation step;
and a detection step of detecting an object based on the detection condition changed by the detection control step and an image captured by the camera.

Images