JP4857072B2 - Monitoring system - Google Patents

Description

  The present invention relates to a monitoring system for monitoring passengers in a car entering a parking lot.

  Conventionally, as a monitoring system for monitoring passengers in a car entering a parking lot, a car is in a parking area where a single car can be parked, and a door opening / closing sound is detected. When it is detected that the engine has been running for a long time, it is determined that there is a person remaining in the vehicle (for example, a child) (for example, see Patent Document 1), and a parking lot is provided with a CCD camera at the entrance of the parking lot A user who enters the camera is imaged and compared with a suspicious person image stored in a suspicious person image database to warn the visit of the suspicious person (for example, see Patent Document 2).

JP 09-171594 (page 3, FIG. 7) JP 2003-216775 A (page 6, FIG. 16)

  However, in the above-described detection device disclosed in Patent Document 1, for example, when the engine is stopped even though the remaining person in the vehicle is a child, the leaving of the child cannot be detected. Even if it is in the unauthorized monitoring system of Patent Document 2 described above, the passenger of the car entering the parking lot is checked against the suspicious person's video, Since it is a warning and does not detect whether the occupant has a child, there is no warning of entry of these children. Therefore, like the detection device of Patent Document 1, the child is left in the car. There was a problem that could be.

  The present invention has been made paying attention to such a problem, and by detecting that a child is present in a car occupant entering a parking lot, the child is left behind in the car. An object is to provide a monitoring system capable of preventing as much as possible.

In order to solve the above-described problem, a monitoring system according to claim 1 of the present invention provides:
A monitoring system for monitoring passengers in a car 1 entering a parking lot,
Imaging means (front camera 2, rear seat camera 4) for capturing an image including at least an in-vehicle image of the automobile 1;
Driver image area specifying means (CPU 102, S3; driver image area specifying process) for specifying a driver's image area from the images picked up by the image pickup means;
Child video determination for determining whether or not there is a video portion with a suspected child in an area other than the driver video area specified by the driver video area specifying means of the video imaged by the imaging means Means (CPU102, S7, S13, S17);
Notification means (CPU 102, S19, hall computer 10) for making a notification based on the determination result in the child video determination means being a determination result that there is a video portion suspected of being a child;
It is characterized by having.
According to this feature, it is determined whether or not there is a video portion that is suspected of being in a region other than the driver's video region in the captured video, and notification is made based on the determination that it is present. Therefore, it becomes possible to recognize that the child is riding with high efficiency and with high accuracy, and it is possible to prevent these children from being left in the vehicle.

The monitoring system according to claim 2 of the present invention includes:
A monitoring system for monitoring passengers in a car 1 entering a parking lot,
Thermograph imaging means (front thermograph camera 2 ′, rear seat thermograph camera 4 ′) for capturing a thermograph image including at least an in-vehicle thermograph image of the automobile 1;
Driver image area specifying means (CPU 102, Sb3; driver image area specifying process) for specifying a driver's image area from among the thermograph images captured by the thermograph imaging means;
Determining whether or not there is a video part that is suspected of a child in an area other than the driver video area specified by the driver video area specifying means of the thermographic video imaged by the thermograph imaging means Children video determination means (CPU102, Sb7, Sb13, Sb17),
Notification means (CPU 102, Sb19, hall computer 10) for making a notification based on the determination result in the child video determination means being a determination result that there is a video portion suspected of being a child;
It is characterized by having.
According to this feature, it is determined whether or not there is an image portion suspected of having a child in an area other than the driver's image area in the captured thermograph image, and based on the determination that it exists. Therefore, it becomes possible to recognize that the child is riding with high efficiency and with high accuracy, and therefore it is possible to prevent the child from being left in the vehicle.

A monitoring system according to claim 3 of the present invention is the monitoring system according to claim 2,
In addition to the thermograph imaging means (front thermograph camera 2 ′, rear seat thermograph camera 4 ′), it includes identification video imaging means (front camera 2) that captures images that can individually identify the automobile,
The notification means (CPU 102, Sb19, Hall computer 10) determines that the determination result (Sb7, Sb13, Sb13, Sb13, Sb17) includes a video portion that is suspected of being a child. It is characterized by including display means (hall computer 10) for displaying an image captured by the identification image capturing means of the vehicle determined to exist based on being Yes at Sb17.
According to this feature, an image of a car that has been determined to contain an image part suspected of having a child will be displayed. Can be easily identified and discovered.

A monitoring system according to claim 4 of the present invention is the monitoring system according to any one of claims 1 to 3,
Child recognition reference data storage means (storage unit 105) for storing child recognition reference data (region area threshold data) serving as a reference for recognizing a child;
The child image determination means (CPU 102, S7, S13, Sb7, Sb13) uses the child recognition reference data stored in the child recognition reference data storage means to use the imaging means (front camera 2, rear seat camera 4). Regions other than the driver's video region (S3 and Sb3 driving) of the image captured by the driver or the thermograph image captured by the thermograph imaging means (front thermograph camera 2 ′, rear seat thermograph camera 4 ′) In-vehicle image areas other than the driver image area specified in the hand image area specifying process (including the in-vehicle image area in the rear seat window frame specified in the in-vehicle image area specifying process in S9 and Sb9 and S15 and Sb15) It is characterized by determining whether or not there is a video part suspected of being a child.
According to this feature, by using the child reference recognition data, it can be recognized that the child is riding with higher accuracy.

A monitoring system according to a fifth aspect of the present invention is the monitoring system according to any one of the first to fourth aspects,
The driver image area specifying means (CPU 102) is an image picked up by the image pickup means (front camera 2, rear seat camera 4) or the thermograph image pickup means (front thermograph camera 2 ′, rear seat thermograph camera). 4 ′), the in-vehicle image area corresponding to the inside of the vehicle is identified (CPU102, S2, S9, S15, Sb2, Sb9, Sb15), and the above-mentioned in-vehicle image area is selected from the specified in-vehicle image area. It is characterized by specifying the image area of the driver (S3, Sb3).
According to this feature, the driver's image area is specified from the in-vehicle image area, so that the driver's image area can be efficiently and quickly compared with the case where the driver's image area is specified from the entire image. Therefore, the processing speed of the entire system can be improved.

A monitoring system according to claim 6 of the present invention is the monitoring system according to any one of claims 1 to 5,
The child image determination means (CPU 102) is a video imaged by the imaging means (front camera 2, rear seat camera 4) or the thermograph imaging means (front thermograph camera 2 ′, rear seat thermograph camera 4 ′). ), The vehicle image area corresponding to the vehicle is identified (CPU102, S2, S9, S15, Sb2, Sb9, Sb15), and the child is suspected from the identified vehicle image area. It is characterized in that it is determined whether or not a certain video portion exists (S7, S13, S17, Sb7, Sb13, Sb17).
According to this feature, since it is determined whether or not there is an image portion that is suspected of being a child from the image in the vehicle image area other than the image area of the driver, the driver in the entire image is determined. Compared to the case where a region other than the video region is a target for determining whether or not there is a video part that is suspected of being a child, the determination can be performed more efficiently and quickly. It can be improved.

A monitoring system according to a seventh aspect of the present invention is the monitoring system according to any one of the first to sixth aspects,
The imaging means or the thermograph imaging means rides on the rear seat of the automobile in addition to the front-side imaging means (the front camera 2 and the front thermograph camera 2 ′) that images the automobile 1 from the front substantially front. It is characterized by having rear seat imaging means (rear seat camera 4, rear seat thermograph camera 4 ′) for imaging a passenger.
According to this feature, it is possible to more reliably detect the child riding in the rear seat by providing the rear seat imaging means exclusively for the rear seat.

A monitoring system according to claim 8 of the present invention is the monitoring system according to claim 7,
The child image determining means determines that there is an image portion suspected of being a child on the condition (Yes in S17) that it is determined that a passenger's image exists in the captured image of the rear seat imaging means. It is characterized by.
According to this feature, since there is a high possibility that a child is riding on the rear seat, it is possible to easily detect that there is a possibility of child riding.

A monitoring system according to claim 9 of the present invention is the monitoring system according to claim 7 or 8,
Whether the child video determination means includes an image of a passenger in an area other than the driver's video area in an image captured by the front side imaging means (front camera 2, front thermograph camera 2 '). (S5, Sb5), and if the determination result is a determination result that there is no passenger image, the rear seat imaging means (rear seat camera 4, rear seat thermograph camera 4 ') On the condition that it is determined that there is an image of the passenger in the captured image (Yes in S17 and Sb17), it is determined that there is an image portion suspected of being a child, while the determination result indicates that the image of the passenger is present. If it is determined that there is a passenger image in the captured image of the rear seat imaging means (Yes in S11 and Sb11), the result of the determination is that there is an image captured by the rear seat imaging means. Occupant images, determines whether the pictures of children suspected (S13, Sb13) is characterized in that.
According to this feature, when it is determined that the passenger's image does not exist in the captured image of the front-side imaging means, that is, when it is determined that the passenger is not in the front seat, imaging for the rear seat is performed. When it is determined that there is a passenger image in the captured image of the means, that is, it is determined that there is a passenger in the rear seat, it is possible to easily carry a child by determining that the passenger is a child. When it is determined that there is an image of the passenger in the captured image of the front-side imaging means, that is, when it is determined that the passenger is in the front seat, By determining whether the image of the passenger in the captured image of the imaging means is a video that is suspected of being a child, it is possible to detect the child's riding more accurately, and thus, practically quickly Processing and accuracy can be obtained

A monitoring system according to claim 10 of the present invention is the monitoring system according to any one of claims 1 to 9,
Including in-field imaging means (in-field cameras 5a to 5d) for capturing images in the parking lot, including the car that has entered the parking lot;
In the in-field video imaged by the in-field imaging means, the car 1 determined by the child video determination means (CPU 102, S7, S13, S17, Sb7, Sb13, Sb17) that there is an image portion suspected of being a child exists. Vehicle identification means (CPU 102, identification display processing program; display target vehicle identification processing) to be identified and child car identification display means (CPU 102, identification display processing program; identification number display processing) for displaying the identified vehicle in an identifiable manner ).
According to this feature, a store clerk or a monitor can easily recognize the location of a car that is determined to have a video part that is suspected of being a child, so that a child can travel with it. It is possible to easily identify and find the car that is likely to be in the parking lot.

  Examples of the present invention will be described below.

  An embodiment corresponding to claim 1 of the present invention will be described with reference to the drawings. First, FIG. 1 is a diagram showing an overview of a monitoring system according to the first embodiment of the present invention.

  As shown in FIG. 1, the parking lot to which the monitoring system of the present embodiment is applied is a substantially square parking lot provided on the back side of the game hall facing the road, and the side of the game hall. In the position, each lane of the entrance to the parking lot is provided.

  And the monitoring system of a present Example is the front camera 2 used as the front side imaging means in this invention provided in the vicinity of the entrance lane of a parking lot so that the image | video in the vehicle 1 of the motor vehicle 1 which enters a parking lot can be imaged. A rear seat camera 4 provided at the upper part of the ticketing machine 3 for issuing a parking ticket, and in-field cameras 5a to 5d serving as in-field imaging means in the present invention provided at the four corners of a substantially square parking lot, The monitoring terminal 100 (see FIG. 3) installed in the amusement hall is mainly configured.

  FIG. 2 is a perspective view showing a state of the entrance lane in the present embodiment. The entrance lane includes an opening / closing bar 6 for restricting the entrance of the automobile 1 to the parking lot, and an opening / closing for opening / closing the opening / closing bar 6. An opening / closing gate 15 comprising the device 9 and a ticketing machine 3 for issuing a parking ticket are provided at a position facing the driver's seat of the automobile 1 parked in the immediate vicinity of the opening / closing gate 15. When the parking ticket is taken out, the opening / closing bar 6 is opened so that the automobile 1 can enter.

  In addition, various devices that are located obliquely forward when viewed from the automobile 1 of the opening / closing bar 6 are installed on the island I, which is one step higher than the roadway, by being pivotally supported by the upper end of the column 12. In the present invention, a front camera 2 that can image the inside of the automobile 1 from the upper right direction of the automobile 1 is installed, and the rear seat of the entering automobile 1 is imaged at the upper position of the ticketing machine 3. A rear seat camera 4 serving as an imaging means for the rear seat is installed. That is, the image pickup means of the present invention is formed by the front camera 2 and the rear seat camera 4 of the present embodiment.

  The front camera 2 and the rear seat camera 4 used in the present embodiment are infrared cameras, and by projecting infrared rays from themselves, it is possible to satisfactorily display an image in the vehicle without depending on ambient brightness. While being able to image and projecting infrared rays that are not visible light for imaging purposes, the passenger is prevented from being aware that the image is being captured by the camera, but the present invention is not limited to this. However, these front camera 2 and rear seat camera 4 may be ordinary visible light cameras, as long as they can take a good image of the interior of the vehicle without depending on ambient brightness, etc. These visible light camera and infrared camera may be provided side by side.

  Further, the front camera 2 and the rear seat camera 4 used in the present embodiment have a vertical tilting function (swing function), and the monitoring terminal 100 to which the front camera 2 and the rear seat camera 4 are connected. Thus, the angle control in the vertical direction is executed.

  In the present embodiment, rod-shaped vehicle height detectors 7a and 7b for detecting the vehicle height of the entering vehicle 1 are provided on both sides of the entrance lane where the vehicle stops.

  In the vehicle height detectors 7a and 7b, a bar-shaped unit in which a plurality of infrared light emitting units and infrared light receiving units are arranged is arranged in the height direction so that the infrared light emitting unit and the infrared light receiving unit face each other. When the vehicle 1 stops between the infrared light emitting unit and the infrared light receiving unit, the vehicle height of the entering vehicle 1 is detected based on the height position of the infrared light receiving unit at the highest position where the infrared light is received. The detection results (signals) of these infrared light receiving units are output to the monitoring terminal 100 connected to the vehicle height detectors 7a and 7b (see FIG. 3). At 100, the monitoring terminal 100 can detect the height of the car 1 that has entered.

  Further, as shown in FIG. 2, a vehicle sensor 8 made of a metal sensor is installed between the vehicle height detectors 7a and 7b, and the vehicle sensor 8 is also shown in FIG. , Connected to the monitoring terminal 100.

  Then, when the vehicle 1 has entered, the vehicle sensor 8 detects the vehicle 1 and outputs a detection signal to the monitoring terminal 100, so that the monitoring terminal 100 detects that the vehicle has entered the entrance lane. It can be detected.

  In addition, the in-field cameras 5a to 5d of the present embodiment are normal visible light fixed cameras, and these in-field cameras 5a to 5d are connected to the monitoring terminal 100 installed in the game hall ( As shown in FIG. 16, the images captured by the on-site cameras 5a to 5d are simultaneously displayed on the display device 107 of the monitoring terminal 100, and the monitoring terminal 100 grasps the situation inside the field. It can be done.

  Reference numeral 14 in FIG. 1 is an open / close gate that is installed at the exit of the parking lot and regulates the participation of automobiles. Reference numeral 13 in FIG. 1 is a settlement apparatus for settlement of the parking fee.

  Next, the monitoring terminal used in the present embodiment will be described. The monitoring terminal is installed in a game hall, for example, in an office, and as shown in FIG. 3, the front camera 2, the rear seat camera 4, The on-site cameras 5a to 5d are connected to the vehicle sensor 8 and the vehicle height detectors 7a and 7b.

  In addition, the monitoring terminal 100 of the present embodiment is also connected to the hall computer 10 that collects and manages information related to gaming of a gaming machine such as a pachinko machine or a slot machine installed in the game hall so that data communication is possible. The entire video data of the automobile 1 determined that the child is on the hall computer 10 can be transmitted to the hall computer 10 for display.

  Here, the configuration of the monitoring terminal 100 used in the present embodiment will be described based on the block of FIG. 4. The monitoring terminal 100 of the present embodiment is formed by a computer, and data in the computer is stored inside the computer. A CPU 102 that performs various processes performed by the monitoring terminal 100 based on a processing program, a RAM 103 that is used as a temporary storage area during processing, and the current time information and calendar information are output to the data bus 101 that performs transmission and reception A real-time clock (RTC) 104, a storage unit 105 including a magnetic disk or a magneto-optical disk, an input device 106 such as a keyboard and a mouse, a display device 107 for displaying and outputting various information, a vehicle sensor 8, and vehicle height detectors 7a and 7b. To the gate control device for controlling the operation of the switchgear 9 An external I / O unit 108 that outputs an end signal, a rear seat camera 4, and in-field cameras 5 a to 5 d are connected, and a video capture unit 109 that captures video based on video signals output from each camera, and the rear seat camera 4 In addition, the on-site cameras 5a to 5d and the communication unit 110 that performs data communication with the hall computer 10 are connected.

  In addition to the processing program in which processing contents such as processing shown in FIG. 15 described later are described, the storage unit 105 includes shape data (front glass frame) of the windshield frame of each vehicle type for specifying the vehicle interior area. Data) and rear frame window frame shape data (rear seat window frame data), wheel shape data for each vehicle model for recognizing the driver's image area, and from almost the front of a person (adult) Driver contour data consisting of human (adult) contour data consisting of looked shape data, and face contour consisting of various face contour shape data used to identify the face contour from the contours extracted from the video The data, the human contour data for specifying the human region corresponding to the face contour, or the region area calculated in FIG. 15 described later is an adult region area or a child And region area threshold data for determining whether the region area is stored. That is, in the present embodiment, the storage unit 105 stores steering wheel shape data and person contour data serving as driver recognition data, and the storage unit 105 forms the driver recognition data storage unit of the present invention. .

  In addition, the storage unit 105 of the present embodiment stores area area threshold data serving as child recognition reference data in the present invention, and the storage unit 105 forms a child recognition reference data storage unit in the present invention. Yes.

  Hereinafter, the operation of the monitoring system of the present embodiment will be described with reference to FIGS. 5 to 16. First, when the vehicle 1 enters the entrance lane and stops immediately before the open / close gate, the vehicle 1 enters the vehicle. Detected by the sensor 8, a detection signal is output to the monitoring terminal 100.

  In response to the input of the detection signal from the vehicle sensor 8, the monitoring terminal 100 specifies the vehicle height of the vehicle 1 that has entered the vehicle based on the detection signals output from the vehicle height detectors 7a and 7b. The angle control data is transmitted to the front camera 2 and the rear seat camera 4 so that the angles of the front camera 2 and the rear seat camera 4 become the angle according to the vehicle height, and the angle control is performed.

  After the angle control of the front camera 2 and the rear seat camera 4 is performed, the camera images captured by the front camera 2 and the rear seat camera 4 such as the images shown in FIGS. 5 and 6 are captured. After the execution, the processing shown in FIG. 15 is started.

  That is, as shown in FIG. 5, the image captured by the front camera 2 is an image captured from the front through the windshield, and the image captured by the rear seat camera 4 is shown in FIG. As shown, the rear seat of the automobile 1 is captured from an oblique front through the rear seat window.

  Specifically, in the processing shown in FIG. 15, among these captured videos, the video of the front camera 2 shown in FIG. 5 is subjected to image processing to perform contour extraction processing (S1). As disclosed in the above-mentioned prior application, specific image processing methods for contour extraction are applied to a well-known image processing method, for example, the image shown in FIG. As shown in FIG. 4, it is only necessary to obtain a contour-extracted video image whose contour has been extracted.

  Then, an in-vehicle image area specifying process for specifying an in-vehicle image area in the contour extracted image of FIG. 7 obtained in step S1 is performed (S2). Specifically, in this in-vehicle image area specifying process, based on the glass frame data of the windshield out of the glass frame data stored in the storage unit 105, included in the windshield frame data in these contour extraction images. The contour matching the shape of the various windshield frames is extracted by pattern matching processing, the extracted contour is specified as the windshield frame, and the video region within the specified contour is specified as the vehicle video region (see FIG. 8).

  As a method of pattern matching processing, a known pattern matching method, for example, a template image, specifically, correlation with windshield frame data of each vehicle type stored in the storage unit 105 is calculated by Fourier transform. Then, the contour having the highest correlation value, that is, the highest matching degree is extracted. At this time, the contour value of the windshield frame data is appropriately fed back so that the contour of the windshield frame data is used by rotating the angle, horizontal and vertical movements, and being enlarged / reduced within a predetermined range. May be processed.

  And it progresses to the step of S3 and implements the driver | operator image | video area | region specific process which pinpoints a driver | operator's image | video area | region for the inside of the vehicle image | video area | region shown in FIG. 8 specified by the step of S2. Specifically, in the driver image area specifying process, based on various handle shape data included in the driver recognition data stored in the storage unit 105 in the in-vehicle image area specified in step S2. As described above, the contour matching the shape is identified as the handle contour as shown in FIG. 9 by performing a known pattern matching process, and then the human contour corresponding to the identified handle contour, That is, as shown in FIG. 8, the image region inside the contour that matches the shape based on the human (adult) contour data that includes the contour of the specified handle is processed in the same manner as the pattern matching processing described above. To identify as a driver image area.

  That is, in the driver image area specifying process of step S3 performed by the CPU 102 of the present embodiment, the present embodiment is performed using the handle shape data and the human contour data that are the driver recognition data stored in the storage unit 105. The driver's image area is identified from among the images captured by the front camera 2 that constitutes the imaging means of the example, and further from the in-vehicle image area specified by the in-vehicle image area specifying process of S2. The driver image area specifying means in the present invention is formed by the CPU 102 that performs the driver image area specifying process.

  Then, the process proceeds to step S4, and is stored in the storage unit 105 as a determination target area, which is an area other than the driver video area specified in step S3 and in the in-vehicle area specified in step S2. Face contour presence / absence determination processing is performed by determining whether or not there is a contour that matches the shape based on the existing face contour data by performing the above-described pattern matching processing.

  If it is determined in S4 that there is no face outline in the determination target area, the process proceeds to step S14, and the rear seat camera video captured along with the front camera video is contoured in the same manner as in S1. Extraction processing is performed to obtain a contour-extracted video image whose contour has been extracted as shown in FIG.

  Then, an in-vehicle image area specifying process for specifying an in-vehicle image area in the contour extracted image of FIG. 11 obtained in step S14 is performed (S15). Specifically, in this in-vehicle image area specifying process, based on the rear seat window frame data in the glass frame data stored in the storage unit 105, these contour extracted images include the rear seat window frame data. The contours that match the shapes of the various rear seat window frames are identified as the rear seat window frames by performing the pattern matching process described above, and the video region within the identified contours is identified as the in-vehicle video region (see FIG. 12).

  Then, the process proceeds to step S16, and whether or not there is a contour that matches the shape based on the face contour data stored in the storage unit 105 in the in-vehicle video area shown in FIG. 12 specified in step S15. The face contour presence / absence determination process is performed by performing the pattern matching process described above.

  If the determination result in S16 is a determination result that there is no face contour in the target area (S17), the process proceeds to step S21, and a completion signal is transmitted to the gate control device to open the gate. The vehicle 1 is allowed to enter the parking lot and the process is terminated. On the other hand, if it is determined that there is a face outline in the target area (S17), the process proceeds to step S18, where the child After determining that there is a video area, after transmitting an instruction to shift to the zoomed-out overall imaging state capable of capturing the entire image including the license plate of the entering car 1 to the front camera 2, the entire image is captured. The current time information output from the real-time clock (RTC) 104 and the identification number by the serial number on that day are stored together with the stored whole image. Against hall computer 10, it transmits along with notification request No. (S19).

  That is, in the step S17 performed by the CPU 102 of the present embodiment, the rear seat serving as the imaging unit of the present embodiment, which is an area other than the driver video area specified in the driver image area specifying process (S3). It is determined whether or not there is a video portion that is suspected of being present in the video area captured by the camera 4, and the child video determination means of the present invention is formed by the CPU 102 that executes step S 17. Yes.

  Further, in the step S19 performed by the CPU 102 of the present embodiment, there is a video portion in which the determination result in the step S17 serving as the child video determination means in the present invention and each of the steps S7 and S13 described later is suspected of being a child Based on the determination result, a notification request is transmitted to the hall computer 10 to notify the hall computer 10, and the CPU 102 that executes the step of S19 and the hall computer 10 perform the notification in the present invention. An informing means is formed.

  The hall computer 10 that has received this notification request displays the identification number and the entire video received together with the notification request, and confirms that there is an entrance to a parking lot of a car that is likely to be accompanied by a child. , Notify the store clerk.

  Then, by setting the vehicle 1 as an identification display target, the vehicle 1 set as the identification display target by the above-described identification display processing program executed in parallel with these processes is displayed in each in-field camera 5a. The vehicle 1 that is specified on the video of -5d and is the identification display target is displayed on the video of the on-site cameras 5a-5d as shown in FIG. .

  Specifically, when a new identification display target is set, the identification display processing program sets an identification line corresponding to the end position of the entrance lane provided on the images of the in-field cameras 5a to 5d. Then, a display target vehicle specifying process for specifying the vehicle 1 that has passed for the first time after the setting of the new identification display target as the identification display target vehicle is performed, and the specified vehicle is moved along with the movement of the specified vehicle. By performing the identification number display process for moving and displaying the identification numbers assigned to 1, the images of the on-site cameras 5a to 5d to which these identification numbers are displayed are displayed on the display device 107 as shown in FIG. Is displayed.

  That is, in the display target vehicle identification process performed by the CPU 102 of the present embodiment, the vehicle in which it is determined that there is a video portion suspected of being a child in each of steps S7, S13, and S17 serving as the child video determination unit of the present invention. 1 is specified, and the vehicle specifying means in the present invention is formed by the CPU 102 that performs the display target vehicle specifying process.

  In the identification number display process performed by the CPU 102 of this embodiment, the vehicle 1 specified in the display target vehicle specifying process is displayed so as to be identifiable, and the CPU 102 that executes the display target vehicle specifying process displays the present invention. The child car identification display means is formed.

  And it progresses to the step of S21, a completion signal is transmitted to a gate control apparatus, a gate is opened, entry into the parking lot of this car 1 is permitted, and this processing is ended.

  That is, in this embodiment, it is determined that there is no face outline in the determination of S4, that is, when there is no passenger other than the driver in the image of the front camera, the face image in the vehicle image area of the rear seat camera When it is determined that there is a contour, that is, when there is a passenger in the rear seat, it is very likely that the passenger in the rear seat is a child seated in a child seat, so the passenger is an adult It is determined that there is a child riding without determining in detail whether the child is a child or not, and the notification is carried out in these situations, that is, the situation of the images shown in FIG. 5 and FIG. In this case, it is determined that there is a child riding, and notification is performed.

  That is, in this embodiment, the CPU 102 serving as the child image determination means of the present invention boarded in the image captured by the front camera 2 (front image capturing means) in the step S5 and other than the image area of the driver. Whether or not there is an image of the passenger, and when the determination result is a determination result that the image of the passenger does not exist, the boarding is included in the captured image of the rear seat camera 4 (imaging means for the rear seat) On the condition that it is determined that there is a person's video, specifically, on the condition that the result of step S17 is Yes, it is determined that there is a video part suspected of having a child, while the determination result is When it is determined that the video is present, and it is determined that the video of the occupant is present in the captured video of the rear seat camera 4, specifically, on the condition that the result of step S11 is Yes, Occupant image in captured image parts seat camera 4, whether the pictures of children suspected, it is determined by implementing the steps of S13.

  On the other hand, in the determination of S5, if it is determined that there is a face outline in the determination target area, that is, if there is a passenger other than the driver, the process proceeds to step S6, and the area corresponding to the existing face outline The area, that is, the outline of the person connected to the face outline is specified by performing the above-described pattern matching process based on the person outline data, and the area area of the specified person outline is calculated.

  Then, the calculated area area is compared with the area area threshold value data stored in the storage unit 105, and it is determined whether or not the area area is equal to or less than the threshold value (S7). That is, in the step S7 performed by the CPU 102 of the present embodiment, the driver identified by the driver image area identifying process (S3) of the image captured by the front camera 2 serving as the imaging means of the present embodiment. It is determined whether or not there is a video part that is suspected of being out of the in-vehicle video area specified in the in-vehicle video area specifying process of S2 and is an area other than the video area, The child image determining means of the present invention is formed by the CPU 102 that executes the steps.

  Further, in this embodiment, the CPU 102 serving as the child image determination unit uses the area camera threshold data serving as the child recognition reference data stored in the storage unit 105 to capture an image using the front camera 2 serving as the imaging unit of the present invention. There is a video portion in which the child is suspected in a region other than the driver video region of the captured video, specifically, in the vehicle video region other than the driver video region specified in the driver video region specifying process of S3. It is determined whether or not.

  If the area is equal to or smaller than the threshold value, it is considered that a child is riding in the passenger seat or the like. Therefore, after proceeding to step S18 and determining that there is a child video area, S19 to S21 described above are performed. After each step is performed, an instruction to shift to a zoomed-out overall imaging state capable of imaging the entire image including the license plate of the entering car 1 is transmitted to the front camera 2, and then the entire image is displayed. Capture and store the current time information output from the real-time clock (RTC) 104 together with the identification number based on the serial number on that day, and notify the hall computer 10 of the stored whole image and identification number. By transmitting with the request (S19), in the hall computer 10, there is a high possibility that the child is on board. Of that there was an entry to the parking lot is notified to the clerk or the like.

  On the other hand, if it is determined in S7 that the area of the area exceeds the threshold value, it is considered that an adult or the like is riding in the passenger seat or the like, so the process proceeds to step S8 and the rear seat camera captured together with the front camera image is taken. Contour extraction processing is performed on the four rear seat camera images in the same manner as in S1 to obtain contour-extracted contour-extracted images as shown in FIG.

  Then, an in-vehicle image area specifying process for specifying an in-vehicle image area in the contour extracted image of FIG. 11 obtained in step S8 is performed (S9). Specifically, in this in-vehicle image area specifying process, based on the rear seat window frame data in the glass frame data stored in the storage unit 105, these contour extracted images include the rear seat window frame data. The contours that match the shapes of the various rear seat window frames are identified as the rear seat window frames by performing the pattern matching process described above, and the video region within the identified contours is identified as the in-vehicle video region (see FIG. 12).

  Then, the process proceeds to step S10, and there is a contour that matches or approximates the shape based on the face contour data stored in the storage unit 105 in the in-vehicle video area shown in FIG. 12 identified in step S9. Whether or not the face contour is determined by performing the above-described pattern matching processing is performed.

  If the determination result in S10 is a determination result that there is no face contour in the target area (S11), the process proceeds to step S21, and a completion signal is transmitted to the gate control device to open the gate. In the meantime, the entry of the car 1 into the parking lot is permitted and the process is terminated. On the other hand, if it is determined that there is a face outline in the target area (S11, see FIG. 13), the process proceeds to step S12. Then, the area area corresponding to the existing face outline, that is, the outline of the person connected to the face outline is specified by performing the pattern matching process based on the person outline data (FIG. 14). The area area of the contour is calculated.

  At this time, when the position of the specified person contour is a position corresponding to the far seat of the rear seat camera 4, the focus information of the position of the person contour is acquired from the rear seat camera 4, By correcting the area of the area based on the focus information, a person existing in the rear seat far from the rear seat camera 4 may not be erroneously determined as a child.

  Then, the calculated area area is compared with the area area threshold value data stored in the storage unit 105, and it is determined whether or not the area area is equal to or less than the threshold value (S13). That is, in the step S13 performed by the CPU 102 of this embodiment, the rear seat serving as the imaging means of this embodiment, which is an area other than the driver's video area specified in the driver image area specifying process (S3). It is determined whether or not there is a video portion that is suspected of being present in the video area imaged by the camera 4, and the child video determination means of the present invention is formed by the CPU 102 that executes step S 13. Yes.

  Further, in the present embodiment, the CPU 102 serving as the child image determination unit uses the region area threshold data serving as the child recognition reference data stored in the storage unit 105 to identify the driver image region identification process in S3. It is determined whether or not there is a video part that is suspected of a child in the in-vehicle image area in the rear seat window frame specified in the in-vehicle image area specifying process of S9, which is an in-vehicle image area other than the driver image area. ing.

  If the area is less than or equal to the threshold value, it is considered that a child is riding in the rear seat. Therefore, after proceeding to step S18 and determining that there is a child video area, the above-described steps S19 to S21 are performed. By executing each step, after transmitting an instruction to shift to the zoomed-out overall imaging state capable of imaging the entire image including the license plate of the entering car 1 to the front camera 2, the entire image is captured. Then, the time information output from the real-time clock (RTC) 104 and the identification number by the serial number on that day are stored, and the stored whole image and identification number are notified to the hall computer 10. It is transmitted together with the request (S19).

  By transmitting the notification request, the hall computer 10 displays the entire image received from the monitoring terminal 100 together with the notification request and the identification number, so that the child parking lot is likely to be on the car. The store clerk is notified that there has been an entrance.

  And by these notifications, a store clerk, a monitor, etc. can recognize the image and identification number of the vehicle, and can easily identify the vehicle 1 in the in-field image and the field in FIG. It becomes possible to easily identify whether or not a child is left behind in the vehicle 1 by identifying the side of the vehicle 1 where the vehicle 1 is parked.

  On the other hand, if the area exceeds the threshold value, it is considered that an adult or the like is riding in the rear seat. Therefore, the process proceeds to step S21 and a completion signal is transmitted to the gate control device to open the gate. Then, the entry of the automobile 1 into the parking lot is permitted, and the process ends.

  Next, a second embodiment corresponding to claim 2 will be described with reference to FIGS. As a feature of the second embodiment, in the monitoring system of the first embodiment, infrared cameras are used as the front camera 2 and the rear seat camera 4, whereas in the second embodiment, these infrared rays are used. The difference is that a thermographic camera capable of outputting a thermographic video is used instead of the camera, and other system configurations are the same as those of the monitoring system of the first embodiment. Is omitted, and the characteristic part of the second embodiment, which is a change from the first embodiment, will be described below. Note that the storage unit 105 of the monitoring terminal 100 according to the second embodiment stores handle shape data and person contour data, which are driver recognition data according to the second embodiment, as in the first embodiment. A driver recognition data storage unit is formed in the storage unit 105 that stores handle shape data and human contour data as hand recognition data.

  FIG. 17 is a perspective view showing a state of an entrance lane in a parking lot to which the monitoring system of the second embodiment is applied, and the entrance lane has a front similar to that of the first embodiment that serves as an identification image capturing unit according to the present invention. Along with the camera 2, a front thermograph camera 2 ′ serving as a front-side imaging means in the present invention for capturing a thermograph image inside the vehicle 1 is juxtaposed at the upper end of the column 12 ′. A rear-seat thermograph camera 4 'serving as a rear-seat imaging means for capturing a rear-seat thermograph image in place of one rear-seat camera 4 is installed at an upper position of the ticketing machine 3. Yes.

  The front thermograph camera 2 ′ and the rear seat thermograph camera 4 ′ can output a known thermograph image, that is, an image colored according to the amount (intensity) of infrared rays emitted from the subject. However, it is sufficient that it has a resolution sufficient to specify the subject, and these thermograph images are thermographic by specifying colors corresponding to the temperature of each part of the subject according to the infrared intensity by software or the like. A high-precision infrared camera having a thermograph function for converting a graph image can be used. Preferably, the thermograph image is used to specify the range of the human body, so that the temperature of these human bodies is 36 ° C. What is necessary is just to be able to output thermographic images in which the temperature range before and after is color-coded with a resolution of 1 ° C. or less.

  Further, in this embodiment, when it is determined that the child is on the same vehicle, the front camera 2 is used in the same manner as in the first embodiment in order to capture the entire image that can individually identify the vehicle 1. Although the present invention is not limited to this, for example, if the front thermograph camera 2 ′ to be used is a camera that can switch and output both the thermograph image and the infrared camera image, Only the front thermograph camera 2 ′ may be installed, or the front camera 2 may be a normal visible light camera other than an infrared camera, or if the entire image is not captured, these The front camera 2 may not be provided.

  Further, the front thermograph camera 2 ′ and the rear seat thermograph camera 4 ′ are connected to the monitoring terminal 100 installed in the game hall, like the front camera 2 and the rear seat camera 4 of the first embodiment. The monitoring terminal 100 can capture each thermograph video output from the front thermograph camera 2 ′ and the rear seat thermograph camera 4 ′ in the video capture unit 109, and the front thermograph camera 2 In addition, the imaging angle control in the vertical direction can be performed by a tilting function (swing function) provided in the rear seat thermograph camera 4.

  Hereinafter, the operation of the monitoring system according to the second embodiment will be described with reference to the flowchart of FIG. 19. First, when the vehicle 1 enters the entrance lane and stops immediately before the open / close gate, the vehicle 1 enters. It is detected by the vehicle sensor 8 and a detection signal is output to the monitoring terminal 100.

  In response to the input of the detection signal from the vehicle sensor 8, the monitoring terminal 100 specifies the vehicle height of the vehicle 1 that has entered the vehicle based on the detection signals output from the vehicle height detectors 7a and 7b. The angle control data is transmitted to the front camera 2, the front thermograph camera 2 ′, and the rear so that the angles of the front camera 2, the front thermograph camera 2 ′, and the rear seat thermograph camera 4 ′ become the angle according to the vehicle height. It transmits to seat thermograph camera 4 'and angle control is implemented.

  And after implementing these angle control, after taking in the thermograph image | video output from the front thermograph camera 2 'and the backseat thermograph camera 4', the process shown in FIG. 19 is started. .

  In the processing of FIG. 19 of the second embodiment, first, the captured thermographic video of the front thermograph camera 2 ′ is subjected to image processing, and a color difference, that is, a temperature difference corresponding to the contour in the first embodiment is detected. A boundary extraction process for extracting a large portion is performed (Sb1). Note that the boundary extracted video subjected to the boundary extraction process is temporarily stored separately from the captured front thermograph camera 2 '.

  The specific image processing method of these boundary extractions is, for example, by scanning each pixel of a video, and if the hue difference between adjacent pixels is equal to or greater than a predetermined threshold, the pixel is white and is less than the threshold. For example, by converting to black, it is possible to obtain a binarized video from which the boundary is extracted, that is, a video similar to the video shown in FIG. 7 in the first embodiment described above. However, the method of extracting these boundaries may be image processing using a special filter for extracting these boundaries.

  Then, an in-vehicle image area specifying process for specifying an in-vehicle image area from the boundary extracted image obtained in step Sb1 is performed (Sb2). Specifically, in the captured thermograph image of the front thermograph camera 2 ′, an area in the range of 20 to 30 ° C. corresponding to the room temperature is provisionally specified as the in-vehicle image area, and the in-vehicle image thus provisionally specified Whether the boundary in the boundary extracted video corresponding to the region matches the shape based on the glass frame data of the windshield among the glass frame data stored in the storage unit 105 or not is determined by the above-described pattern matching processing by Fourier transform By performing the above, a matching boundary (correlation value equal to or greater than a predetermined value) is specified as a windshield frame, and an image area within the specified boundary is specified as an in-vehicle image area.

  As described above, in the second embodiment, after the vehicle interior image area is provisionally specified in the range corresponding to the room temperature in the thermographic image, the vehicle interior image area is specified by pattern matching with the glass frame data of the windshield. However, this is preferable because the in-vehicle image area can be specified quickly and with higher accuracy. However, the present invention is not limited to this, and the in-vehicle image can be obtained using only the thermograph image. The area may be specified, or conversely, as in the first embodiment, the in-vehicle image area may be specified using only the boundary extracted image.

  And it progresses to the step of Sb3 and implements the driver | operator image area | region identification process which pinpoints a driver | operator's image | video area | region for the inside of the image | video area | region specified at the step of Sb2. Specifically, as in the driver image area specifying process in the first embodiment, the driver recognition data stored in the storage unit 105 is stored in the in-vehicle image area in the boundary extracted image specified in step Sb2. As described above, the boundary that matches the shape based on the various handle shape data included is identified as the handle outline by performing a known pattern matching process (see FIG. 9). A human body temperature range corresponding to the contour, that is, a video region having a color corresponding to a human body temperature of 35 to 37 ° C. including the contour of the identified handle is identified as a driver video region.

  That is, in the driver image area specifying process in step Sb3 performed by the CPU 102 of the present embodiment, the imaging means of the present embodiment is used by using the handle shape data serving as the driver recognition data stored in the storage unit 105. The image area of the driver is identified from among the images captured by the front thermograph camera 2 ′ forming the vehicle interior, and from the in-vehicle image area specified by the in-vehicle image area specifying process of Sb2. The driver image area specifying means in the present invention is formed by the CPU 102 that performs the driver image area specifying process.

  Then, the process proceeds to the step Sb4, and the region other than the driver image region specified in the step Sb3 and the image region on the thermograph image in the in-vehicle region specified in the step Sb2 is set as the determination target region, the human body temperature The human body part presence / absence determination process for determining whether or not there is a color region corresponding to 35 to 37 ° C. is performed.

  In the determination of Sb4, if it is determined that there is no human body part in the determination target area, the process proceeds to step Sb14 and the thermography of the rear seat thermograph camera 4 ′ captured together with the thermograph video of the front thermograph camera 2 ′. A boundary extraction process is performed on the graph image in the same manner as Sb1 to obtain a boundary-extracted boundary-extracted image similar to the contour image of FIG. 11 in the first embodiment.

  Then, similarly to Sb2 described above, in-vehicle image area specifying processing for specifying the in-vehicle image area is performed (Sb15). That is, in the captured thermograph image of the rear seat thermograph camera 4 ′, an area in the range of 20 to 30 ° C. corresponding to the room temperature is provisionally specified as an in-vehicle image area, and the provisionally specified in-vehicle image area is The above-described pattern matching process is performed by Fourier transform to determine whether the boundary in the corresponding boundary extracted video matches the shape based on the rear seat window frame data in the glass frame data stored in the storage unit 105. Thus, the matching boundary (correlation value is a predetermined value or more) is specified as the rear seat window frame, and the video area within the specified boundary is specified as the in-vehicle video area.

  And it progresses to the step of Sb16, and the area | region of the color applicable to 35-37 degreeC which is a human body temperature exists similarly to Sb4 for the inside in the vehicle image area (on thermograph image) specified by the step of Sb15. A human body part presence / absence determination process for determining whether or not to perform is performed.

  When the determination result of Sb16 is a determination result that there is no human body part in the determination target region (Sb17), the process proceeds to step Sb21, and a completion signal is transmitted to the gate control device to When the result of the determination is that the human body part is present in the determination target area (Sb17), the process proceeds to step Sb18. After determining that there is a child video area, the entire video output from the front camera 2 is captured and stored together with the current time information output from the real-time clock (RTC) 104 and the identification number by the serial number on that day. At the same time, the stored whole image and identification number are transmitted to the hall computer 10 together with a notification request (Sb19).

  That is, in the step of Sb17 performed by the CPU 102 of the second embodiment, the image pickup unit of the second embodiment becomes an area other than the driver's video area specified in the driver video area specifying process (Sb3). It is determined whether or not there is a video portion that is suspected of being present in the video area captured by the rear seat thermograph camera 4 ′, and the CPU 102 that executes the step of Sb17 determines the child video determination of the present invention. Means are formed.

  In addition, in the step Sb19 performed by the CPU 102 of the second embodiment, there is a video portion in which the determination result in the step Sb17 serving as the child video determination unit in the present invention and each step Sb7 and Sb13 described later is suspected of being a child. A notification request is transmitted to the hall computer 10 for the entire video imaged by the front camera 2 of the automobile 1 serving as the identification video imaging means of the present invention determined to exist based on the determination result that it exists. Thus, the information is displayed on the hall computer 10 for notification, and the CPU 102 for executing the step of Sb19 and the hall computer 10 form the notification means in the present invention.

  The hall computer 10 that has received this notification request displays the identification number and the entire video received together with the notification request, and confirms that there is an entrance to a parking lot of a car that is likely to be accompanied by a child. , Notify the store clerk.

  Then, by setting the vehicle 1 as an identification display target, the vehicle 1 that is set as the identification display target by the above-described identification display processing program that is executed in parallel with these processes, as in the first embodiment. Are displayed on the images of the in-field cameras 5a to 5d as shown in FIG.

  And it progresses to the step of S21, a completion signal is transmitted to a gate control apparatus, a gate is opened, entry into the parking lot of this car 1 is permitted, and this processing is ended.

  That is, in the second embodiment, similarly to the first embodiment, it is determined that there is no human body part in the determination of Sb4, that is, when there is no passenger other than the driver in the thermograph image of the front thermograph camera 2 ′. Thus, when it is determined that there is a human body part in the in-vehicle image area of the thermographic image of the rear seat thermograph camera 4 ′, that is, when there is a passenger in the rear seat, the passenger in the rear seat becomes a child seat. Since the possibility of being a seated child is very high, it is determined that there is a child rider and notification is carried out without determining in detail whether the passenger is an adult or a child. ing.

  On the other hand, in the determination of Sb5, when it is determined that there is a human body part in the determination target area, that is, when there is a passenger other than the driver, the process proceeds to step Sb6, and the area corresponding to the existing human body part The area is calculated, and the calculated area is compared with the area area threshold data stored in the storage unit 105 to determine whether the area is equal to or smaller than the threshold (Sb7).

  That is, in the step Sb7 performed by the CPU 102 of the second embodiment, it is specified by the driver video area specifying process (Sb3) of the video imaged by the front thermograph camera 2 ′ serving as the imaging means of the present embodiment. It is determined whether or not there is a video part that is suspected of being out of the in-vehicle image area specified in the in-vehicle image area specifying process of Sb2 and other than the driver's image area. The child image determination means of the present invention is formed by the CPU 102 that executes the step of Sb7.

  Further, the CPU 102 serving as the child image determination unit in the second embodiment uses the area area threshold value data serving as the child recognition reference data stored in the storage unit 105, and the front thermograph camera 2 serving as the imaging unit of the present invention. In the area other than the driver image area of the image picked up in ', specifically in the in-vehicle image area other than the driver image area specified in the driver image area specifying process of Sb3, an image in which the child is suspected It is determined whether or not a part exists.

  If the area is equal to or smaller than the threshold value, it is considered that a child is riding in the passenger seat or the like. Therefore, after proceeding to step Sb18 and determining that there is a child video area, the above-described Sb19 to Sb21 are performed. By executing each step, the entire video output from the front camera 2 is captured and stored together with the time information at that time output from the real-time clock (RTC) 104 and the identification number by the serial number on that day. At the same time, the stored whole image and the identification number are transmitted to the hall computer 10 together with the notification request (Sb19), so that the hall computer 10 can go to the car parking lot where the child is likely to be on board. The store clerk is notified that there has been an entrance.

  On the other hand, in the determination of Sb7, if the area of the area exceeds the threshold value, it is considered that an adult or the like is riding in the passenger seat, etc., so the process proceeds to step Sb8 and the thermograph of the front thermograph camera 2 ′. For the thermograph video of the rear seat thermograph camera 4 ′ captured together with the video, the boundary extraction processing is performed in the same manner as Sb1 to obtain the boundary extracted video with the boundary extracted similar to the contour extracted video shown in FIG.

  Then, in the boundary extracted video obtained in step Sb8, in-vehicle image area specifying processing for specifying the in-vehicle image area is performed (Sb9). That is, in the captured thermograph image of the rear seat thermograph camera 4 ′, an area in the range of 20 to 30 ° C. corresponding to the room temperature is provisionally specified as an in-vehicle image area, and the provisionally specified in-vehicle image area is The above-described pattern matching process is performed by Fourier transform to determine whether the boundary in the corresponding boundary extracted video matches the shape based on the rear seat window frame data in the glass frame data stored in the storage unit 105. Thus, the matching boundary (correlation value is a predetermined value or more) is specified as the rear seat window frame, and the video area within the specified contour is specified as the in-vehicle video area.

  And it progresses to the step of Sb10, and the area | region of the color applicable to 35-37 degreeC which is a human body temperature exists similarly to Sb4 in the vehicle image area (on thermograph image) specified by the step of Sb9. A human body part presence / absence determination process for determining whether or not to perform is performed (Sb11).

  When the determination result of Sb10 is a determination result that there is no human body part in the determination target region (Sb11), the process proceeds to step Sb21, and a completion signal is transmitted to the gate control device, and the gate is turned on. When the result of the determination is that the human body part is present in the determination target area (Sb11), the process proceeds to step Sb12. The area area corresponding to the existing human body part is calculated, and the calculated area area is compared with the area area threshold value data stored in the storage unit 105 to determine whether the area area is equal to or less than the threshold value. (Sb13).

  That is, in the step of Sb13 performed by the CPU 102 of the present embodiment, the rear seat serving as the imaging means of the present embodiment, which is an area other than the driver video area specified in the driver image area specifying process (Sb3). It is determined whether or not there is a video portion that is suspected of being present in the video area captured by the thermograph camera 4 ′, and the CPU 102 that executes the step of Sb13 enables the child video determination means of the present invention. Is formed.

  Further, in the present embodiment, the CPU 102 serving as the child image determination means uses the region area threshold value data serving as the child recognition reference data stored in the storage unit 105 to identify the driver image region identification process in Sb3. It is determined whether or not there is a suspicious image portion of the child in the in-vehicle image area in the rear seat window frame specified in the in-vehicle image area specifying process of Sb9 which is an in-vehicle image area other than the driver image area. ing.

  That is, in the second embodiment, the CPU 102 serving as the child video determination unit of the present invention is in the captured image of the front thermograph camera 2 ′ (front-side imaging unit) in the step Sb5 and is outside the driver's video region. It is determined whether or not a passenger's image is present in the area, and if the determination result is a determination result that there is no passenger's image, the rear seat thermograph camera 4 ′ (rear seat imaging means) On the condition that it is determined that there is an image of the passenger in the captured image, specifically, on the condition that the result of step Sb17 is Yes, it is determined that there is an image part that is suspected of being a child, The determination result is a determination result that there is an image of the passenger, and it is determined that the image of the passenger exists in the captured image of the rear seat thermograph camera 4 ′ (the imaging means for the rear seat). Specifically, on the condition that the result of Sb11 is Yes, the image of the passenger in the captured image of the rear seat thermograph camera 4 '(rear seat imaging means) Whether or not the image is a certain image is determined by performing the step of Sb13.

  If the area is equal to or smaller than the threshold value, it is considered that a child is riding in the rear seat. Therefore, after proceeding to step Sb18 and determining that there is a child video area, each step of Sb19 to Sb21 described above is performed. Is implemented, and the entire video output from the front camera 2 is captured and stored together with the current time information output from the real-time clock (RTC) 104 and the identification number by the serial number on that day, The stored whole image and identification number are transmitted to the hall computer 10 together with a notification request (Sb19).

  By transmitting the notification request, the hall computer 10 displays the entire image received from the monitoring terminal 100 together with the notification request and the identification number, so that the child parking lot is likely to be on the car. The store clerk is notified that there has been an entrance.

  And by these notifications, a store clerk, a monitor, etc. can recognize the image and identification number of the vehicle, and can easily identify the vehicle 1 in the in-field image and the field in FIG. It becomes possible to easily identify whether or not a child is left behind in the vehicle 1 by identifying the side of the vehicle 1 where the vehicle 1 is parked.

  On the other hand, if the area exceeds the threshold value, it is considered that an adult or the like is riding in the rear seat. Therefore, the process proceeds to step S21 and a completion signal is transmitted to the gate control device to open the gate. Then, the entry of the automobile 1 into the parking lot is permitted, and the process ends.

  As described above, according to the second embodiment, since the entire image of the automobile determined to include the image portion suspected of being present is displayed, the vehicle having a high possibility that the child is on board, A store clerk or monitor can easily identify and discover.

  Further, according to each of the above embodiments, it is determined whether or not there is a video part that is suspected of being present in an area other than the driver's video area in the captured video, and it is determined that it exists. The notification will be made based on the above, so that it becomes possible to recognize that the child is on the vehicle efficiently and with high accuracy. Therefore, by leaving these children in the car, these left children It is possible to prevent a fatal accident caused by heat stroke.

  Further, according to each of the embodiments, it is possible to recognize that a child is riding with higher accuracy by using the child recognition data.

  In addition, according to each of the embodiments, the driver's image area is specified from the in-vehicle image area, so that driving can be performed more efficiently and quickly than when the driver's image area is specified from the entire image. The image area of the hand can be specified, and therefore the processing speed of the entire system can be improved.

  Further, according to each of the above embodiments, since it is determined whether or not there is a video part that is suspected of a child from the video in the vehicle video area other than the video area of the driver, Compared with a case where an area other than the driver's video area is a target for determining whether or not there is a video part that is suspected of being a child, the determination can be performed more efficiently and quickly. The processing speed can be improved.

  In addition, according to each of the above embodiments, by providing a rear seat camera (rear seat imaging means) which is a rear seat imaging means dedicated to the rear seat, it is possible to more reliably detect a child riding in the rear seat. can do.

  Further, according to each of the above embodiments, since there is a high possibility that a child is riding on the rear seat, it is possible to easily detect that there is a possibility that the child will ride.

  Further, according to each of the embodiments, when it is determined that no image of the passenger is present in the captured images of the front camera 2 and the front thermograph camera 2 ′ serving as the front-side imaging means, that is, riding in the front seat. When it is determined that the passenger is not on board, there is an image of the passenger in the image taken by the rear seat camera 4 or the rear seat thermograph camera 4 ′ serving as the imaging means for the rear seat, that is, riding on the rear seat. When it is determined that there is a person, it is possible to easily detect that there is a possibility of the child riding by determining that the passenger is a child, and the front camera 2 or the front thermograph camera 2 ′ (front view) When it is determined that the passenger's image is present in the captured image of the side imaging means), that is, when it is determined that the passenger is in the front seat, the rear seat camera 4 and the rear seat thermograph By determining whether or not the image of the passenger in the image captured by the camera 4 ′ (the rear seat imaging means) is an image that is suspected of being a child, it is possible to detect the child riding more accurately. Therefore, practically quick processing and accuracy can be obtained.

  In addition, according to each of the embodiments, a store clerk, a monitor, and the like can easily recognize the location such as where an automobile determined to have a video part suspected of having a child is parked, It is possible to easily identify and find the car in which the child is likely to be on board in the parking lot.

  Although the embodiments of the present invention have been described with reference to the drawings, the specific configuration is not limited to these embodiments, and modifications and additions within the scope of the present invention are included in the present invention. It is.

  For example, in the above-described embodiment, when it is determined that there is a child's video area, the entire image of the automobile 1 is captured in S19. However, the present invention is not limited to this, and the whole of these images is taken. The entire video is always captured in advance for all the cars 1 that enter, and when it is determined that there is a child video area, the entire captured image and the identification number are transmitted to the hall computer 10. You may make it do.

  Moreover, in the said Example, although control, such as zoom-up and zoom-down of the front camera 2 and the rear seat camera 4, is not implemented, this invention is not limited to this, These zoom-up, zoom-down, etc. Is controlled together with tilting control according to the vehicle height, and the windshield frame data and rear seat window frame data stored in the storage unit 105 and the shape of the steering wheel are stored in accordance with the degree of zoom-up and zoom-down. You may make it utilize, correcting the magnification of various data, such as data.

  Moreover, in the said Example, the vehicle height of the motor vehicle 1 to enter is measured, and the tilting control according to vehicle height is implemented, and this makes various motor vehicles 1 with different vehicle heights. However, the present invention is not limited to this, and the image pickup camera has a sufficient resolution and the image pickup lens can use a wide-angle lens. If an imaging area corresponding to various types of automobiles can be secured, the tilting control may not be performed.

  Further, in the above embodiment, the driver's image area is specified using the steering wheel shape data of each vehicle type as the driver recognition data, but the present invention is not limited to this, Without using the steering wheel shape data, if there is one person (face) contour or one person temperature range image area (boundary) in the in-vehicle image area specified in step S2 or Sb2, The one video area may be specified as a driver video area, that is, one driver, and in this case, the driver recognition data corresponds to human or facial contour data. .

  In addition, when there are a plurality of human (face) contours or human temperature range video regions (boundaries) in the in-vehicle video region specified in step S2 or Sb2, the area within each human contour is calculated. In comparison, the region of the human contour having the largest area may be specified as the driver image region. In this case, first, in order to compare the areas, the contour data of the person or the face is used. The human (face) contour or the video region (boundary) of the human temperature range is specified using the Data, etc. will be applicable. In this case, when the area of the passenger seat person is larger, the passenger in the passenger seat may be identified as the driver, but the person (face) contour as the driver recognition data The driver image area is specified based on the data and the comparison rule. Even in this case, it is included in the driver image area specifying means of the present invention.

  In addition, when specifying these driver image areas, a method other than the comparison based on the above-described area, for example, specifying the face contour, the specified face contour, and the windshield specified in the steps S2 and Sb2. The image area with the shorter distance from the frame may be specified as the driver image area. In this case, the driver recognition data includes the contour data of the person and the face, and the windshield frame. This includes program data such as distance comparison rules. In this case, when the distance between the face contour of the passenger seat and the windshield frame is short, it may occur that the passenger in the passenger seat is identified as the driver. The driver image area is specified based on the person (face) contour data and the comparison rule, and even in this case, it is included in the driver image area specifying means of the present invention.

  Moreover, in the said Example, the hall computer 10 displays the whole image | video and identification number of the motor vehicle 1 with which the child may be riding, and parking of the motor vehicle 1 with which the child may be riding. Although it is informed that the vehicle has entered the parking lot, the present invention is not limited to this, and these notifications are also displayed on the display device 107 of the monitoring terminal 100 for implementation. Alternatively, it may be further displayed on another display device, for example, a display installed in a monitoring room or the like, for notification.

  In the second embodiment, only the rear seat thermograph camera 4 ′ is used. However, the present invention is not limited to this, and the rear camera thermograph camera 4 ′ and the front camera 2 and the front thermograph are used. Similarly to the graph camera 2 ′, the rear seat camera 4 used in the first embodiment may be arranged side by side.

  In addition to the thermograph cameras 2 ′ and 4 ′, when the front camera 2 and the rear seat camera 4 ′ that are infrared cameras are also provided, the boundary between the thermograph image including the color corresponding to each temperature and the infrared camera You may make it specify each image area | region of a vehicle interior, a driver | operator, and a boarding passenger using both the outlines based on an image | video.

  In the second embodiment, the in-vehicle image area is specified by performing the in-vehicle image area specifying process in steps S2, Sb2, S9, Sb9, and S15 and Sb15, and the specified in-vehicle image area is determined by the driver. The target area of the video area, the face contour of the passengers, and the human temperature area are set as the target area, and this reduces the recognition error compared to the case where the entire captured video area is the target area. In addition, the processing load for pattern matching processing can be reduced. However, the present invention is not limited to this, and the entire video area is set as the target area without performing the in-vehicle video area specifying process. May be.

  Moreover, in the said Example, as shown in FIG. 16, in the monitoring terminal 100, the motor vehicle determined that the image part with a suspected child exists on the image | video of in-field camera 5a-5d was identified, and this identification was carried out. Although the identification number assigned to the automobile is displayed, the present invention is not limited to this, and the in-field cameras 5a to 5d are not provided, but the identification numbers on the images of these in-field cameras 5a to 5d are displayed. The display may not be performed.

  In the second embodiment, the front camera 2 using an infrared camera is used as the identification video imaging unit. However, the present invention is not limited to this, and the identification video imaging unit may be an automobile individually. Any video that can pick up an identifiable video, for example, a normal visible light video or the like may be used, and a normal camera that can pick up the normal visible light video or the like can also be used.

  In each of the above embodiments, the area area threshold data is used as the child recognition reference data to determine whether or not there is a video portion suspected of being a child. However, the present invention is not limited to this. Instead, as these child recognition reference data, for example, as shown in FIG. 14, child contour data is used, and pattern matching processing with the child contour data is performed, so that the child is suspected. It may be determined whether or not a video portion exists.

  In addition, when performing pattern matching processing based on contours in this way, the child's face contour data is used as child recognition reference data instead of the human contour, and the child's face contour data is used in the pattern matching processing with the child's face contour data. It may be determined whether or not there is a video part that is suspected of.

  Further, a threshold value of the distance from the ceiling to the head of the car 1 is stored as child recognition reference data, and if the distance between the passenger's face contour and the ceiling is equal to or greater than the stored threshold value, the child is suspected. It may be determined that a certain video portion exists.

It is a figure which shows the parking lot to which the monitoring system in Example 1 of this invention was applied. It is a perspective view which shows the mode of the entrance lane of the parking lot to which the monitoring system in Example 1 of this invention was applied. It is a figure which shows the structure of the monitoring system of Example 1 of this invention. It is a block diagram which shows the structure of the monitoring terminal 100 used for the monitoring system of Example 1 of this invention. It is a figure which shows the front camera image | video in the monitoring system of Example 1 of this invention. It is a figure which shows the rear seat camera image | video in the monitoring system of Example 1 of this invention. It is a figure which shows the front camera image | video by which the outline extraction was carried out in the monitoring system of Example 1 of this invention. It is a figure which shows the image | video of the vehicle interior area identified by the vehicle interior area identification process in the monitoring system of Example 1 of this invention. It is a figure which shows the outline of the handle | steering_wheel identified in the driver | operator image area specific process of the monitoring system of Example 1 of this invention. It is a figure which shows the driver | operator image area | region identified in the driver | operator image | video area | region specific process of the monitoring system of Example 1 of this invention. It is a figure which shows the back seat camera image by which the outline extraction was carried out in the monitoring system of Example 1 of this invention. It is a figure which shows the part seat camera image | video of the vehicle interior area | region identified by the vehicle interior area | region identification process in the monitoring system of Example 1 of this invention. It is a figure which shows the face outline specified by the rear seat camera image | video of the monitoring system of Example 1 of this invention. It is a figure which shows the person outline applicable to the face outline specified by the rear seat camera image | video of the monitoring system of Example 1 of this invention. It is a flowchart which shows the processing content implemented in the monitoring terminal 100 of the monitoring system of Example 1 of this invention. It is the image | video of the camera in a field displayed on the monitoring terminal 100 of the monitoring system of Example 1 of this invention. It is a perspective view which shows the mode of the entrance lane of the parking lot to which the monitoring system in Example 2 of this invention was applied. It is a figure which shows the structure of the monitoring system of Example 2 of this invention. It is a flowchart which shows the processing content implemented in the monitoring terminal 100 of the monitoring system of Example 2 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Car 2 Front camera 2 'Front thermograph camera 3 Ticketing machine 4 Rear seat camera 4' Rear seat thermograph camera 5a-5b Indoor camera 6 Opening / closing bar 7a, 7b Vehicle height detector 8 Vehicle sensor 9 Opening / closing device 10 Hall computer 15 Open / close gate 100 Monitoring terminal 101 Data bus 105 Storage unit 106 Input device 107 Display device 108 External I / O unit 109 Video capture unit 110 Communication unit

Claims (10)

A monitoring system for monitoring passengers in a car entering a parking lot,
Imaging means for capturing an image including at least an in-vehicle image of the automobile;
Driver image area specifying means for specifying a driver's image area from the images captured by the imaging means;
Child video determination for determining whether or not there is a video portion with a suspected child in an area other than the driver video area specified by the driver video area specifying means of the video imaged by the imaging means Means,
A notification means for performing notification based on the determination result in the child video determination means being a determination result that there is a video portion suspected of being a child;
A monitoring system comprising: A monitoring system for monitoring passengers in a car entering a parking lot,
Thermograph imaging means for imaging a thermograph image including at least an in-vehicle thermograph image of the automobile;
Driver image area specifying means for specifying a driver's image area from among thermograph images captured by the thermograph imaging means;
Determining whether or not there is a video part that is suspected of a child in an area other than the driver video area specified by the driver video area specifying means of the thermographic video imaged by the thermograph imaging means A child image determination means for
A notification means for performing notification based on the determination result in the child video determination means being a determination result that there is a video portion suspected of being a child;
A monitoring system comprising: In addition to the thermograph imaging means, it comprises identification video imaging means for imaging video that can individually identify the automobile,
The notification means is picked up by the identification image pickup means of the car determined to exist based on the determination result in the child image determination means being a determination result that there is a video portion suspected of being a child. The monitoring system according to claim 2, further comprising display means for displaying the recorded video. Comprising child recognition reference data storage means for storing child recognition reference data serving as a reference for recognizing a child;
The child image determination means uses the child recognition reference data stored in the child recognition reference data storage means, and the driving of the image captured by the imaging means or the thermograph image captured by the thermograph imaging means The monitoring system according to any one of claims 1 to 3, wherein it is determined whether or not there is a video portion suspected of being in a region other than the video region of the hand.   The driver image area specifying means specifies an in-vehicle image area corresponding to the inside of the image taken by the image pickup means or the thermograph image picked up by the thermograph image pickup means, and the specified inside of the vehicle The monitoring system according to claim 1, wherein the driver's video area is specified from the video area.   The child video determination means specifies an in-vehicle image area corresponding to the inside of the image captured by the imaging means or the thermograph image captured by the thermograph imaging means, and the specified in-vehicle image area 6. The monitoring system according to claim 1, wherein it is determined whether or not there is a video portion suspected of being a child.   The imaging means or the thermograph imaging means includes a rear seat imaging means for imaging a passenger in the rear seat of the automobile, in addition to a front-side imaging means for imaging the automobile from the front substantially front. The monitoring system according to any one of claims 1 to 6.   The child image determining means determines that there is an image portion suspected of having a child on the condition that it is determined that an image of a passenger is present in the captured image of the rear seat imaging means. Item 8. The monitoring system according to Item 7.   The child image determining means determines whether or not a passenger's image exists in an area other than the driver's image area within the image captured by the front-side imaging means, and the determination result is the passenger's image. If the result of the determination is that there is no video, it is determined that there is a video portion suspected of being present in the captured image of the rear seat imaging means, provided that it is determined that a video of the passenger exists. On the other hand, when the determination result is a determination result that the passenger's image is present, and it is determined that the passenger's image is present in the captured image of the rear seat imaging means, the rear seat imaging means The monitoring system according to claim 7 or 8, wherein it is determined whether or not an image of a passenger in the captured image is an image suspected of a child. Including in-field imaging means for capturing images in the parking lot, including the car that has entered the parking lot;
In the in-field video imaged by the in-field imaging means, the vehicle specifying means for specifying the car determined by the child video determining means that there is a video portion suspected of being present, and the specified car are displayed in an identifiable manner. The monitoring system according to any one of claims 1 to 9, further comprising: a child car identification display means.

Images