JP5771039B2 - Abandoned person detection device - Google Patents

Description

  The present invention relates to an apparatus for detecting the state of a person existing in a monitoring space by analyzing the captured monitoring image, and particularly in a state where a person who has been deprived of physical freedom is left in the monitoring space. The present invention relates to an abandoned person detection device for detection.

  Conventionally, there is a demand for detecting various crimes from surveillance camera images. For example, when a burglar breaks into a manned store and tries to seize money stored in a safe, he or she may attack the employee because of fear of reporting to the outside or resistance. In some cases, restraining the limbs with a rope or adhesive tape may deprive the body of freedom.

  Victims such as employees who have been deprived of physical freedom in this way will be left in the store after the crime, so victims who were delayed in the detection of robbery incidents due to delayed reporting or who were injured There are concerns about delays in rescue.

  For this reason, the proposal of the abnormality detection apparatus which detects automatically that the victim is in the state which cannot perform the above notification operations, and reports is desired.

  In Patent Document 1, when a monitoring image is analyzed, and the position of the center of gravity of a person existing in the monitoring space does not change for a predetermined time or more, an elderly person who lives alone has become unable to move due to a sudden illness or the like. A monitoring device that automatically detects is disclosed.

JP 2002-44645 A

  Here, there are various cases in which the victim is left in the surveillance space, for example, when the victim is restrained by the intrusion robbery. For example, an abandoned victim may attempt to move, for example, while he is deprived of his physical freedom. In this case, since a change occurs in the position of the center of gravity of the person, it is difficult to detect an abandoned person with the method of detecting that no change occurs in the position of the center of gravity. Therefore, there is a demand for detecting that a victim is left unattended in response to such a situation.

  An object of the present invention is to detect an abandoned person by capturing an image change unique to the abandoned victim.

An abandoned person detection device according to the present invention includes an image acquisition unit that sequentially acquires a monitoring image obtained by photographing a monitoring area, a storage unit that stores a monitoring image in a state where no person is present in the monitoring area, and the image An abandoned person detection apparatus including a control unit that processes a monitoring image acquired by an acquisition unit, wherein the control unit extracts a person region from the monitoring image and the background image, and the person An inverted person detection means for detecting an inverted person area that is in a tilted position in the area, and temporally tracking the inverted person area and in a stationary state in which the person area is not substantially moved or a stationary state determination means for determining certain, the temporally tracking the person area of倒位state, a very low speed movement when the person area is determined not to be the stationary state, and have a very low speed moving below a predetermined speed State And determining a very low speed movement determining means, based on said both determination result of the stationary state determination means and the slow-speed movement determining means, to have a person standing condition detecting means for detecting a state in which the person is left according to a predetermined rule It is characterized by.

In addition, an image acquisition unit that sequentially acquires a monitoring image obtained by photographing the monitoring area via a fisheye lens from above the monitoring area, and a monitoring image in a state where no person is present in the monitoring area are stored as a background image, and a long diameter and a short diameter are stored. An abandoned person detection device comprising: a storage unit that stores a slender figure having a diameter as a person model corresponding to the outer shape of a person; and a control unit that processes a monitoring image acquired by the image acquisition unit. The section compares the person area and the person model with a person area extracting unit that extracts a person area from the monitoring image and the background image, and the major axis of the person area is orthogonal to the radial direction from the center of the image. An inverted person detecting means for detecting an inverted person area that is in an attitude that has fallen when it matches a person model facing the direction, and temporally tracking the inverted person area; A stationary state determination means for determining whether the area is in a stationary state in which the area is not substantially moved, and a slow movement in which the person area in the inverted state is temporally tracked and the person area is moving at a low speed of a predetermined speed or less. A slow movement determination unit that determines a state, and a person leaving state detection unit that detects a state in which the person is left based on the determination results of both the stationary state determination unit and the slow movement determination unit. And

  Further, the person leaving state detecting means is configured such that the elapsed time in the stationary state exceeds a first predetermined value or the elapsed time in the slow movement state exceeds a second predetermined value that is longer than the first predetermined value. In this case, it is preferable to detect that the person is left unattended.

  According to the present invention, it is possible to detect a person who has been left in a state where a reporting operation cannot be performed due to the freedom of the body by a robber, and it is possible to report such an abnormal situation.

It is the schematic diagram which showed the structure of the image of the monitoring space, and the abandoned person detection apparatus. It is a block diagram which shows the structure of an abandoned person detection apparatus. It is a figure which shows a standing model among the person models according to attitude | position. It is a figure which shows an inversion model among the person models according to attitude | position. It is a flowchart which shows the operation | movement of abnormality detection.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

"overall structure"
A report system 1 that includes an abandoned person detection device according to the present embodiment, detects an abnormal scene in which an abandoned person exists, and notifies a security center or the like of the abnormal scene will be described.

  FIG. 1 is a schematic diagram showing an image of the configuration and arrangement of the notification system 1.

  The notification system 1 includes an abandoned person detection device 2, a controller 3, and a center device 5.

  The abandoned person detection device 2 is installed on the ceiling of a room in which a room in which an important object such as a safe 6 is installed is used as a monitoring space. The abandoned person detection device 2 outputs an abnormal signal when detecting an abnormal scene in which the abandoned person exists in the monitoring space.

  The abandoned person detection device 2 is connected to a controller 3 via a communication line, and the controller 3 is connected to a center device 5 such as a security center installed in a remote place via a wide area communication network 4 such as a telephone line or an internet line. Is done. The abnormal signal output from the abandoned person detection device 2 is transmitted to the center device 5 via the controller 3.

“Configuration of the abandoned person detection device”
FIG. 2 is a block diagram showing the configuration of the abandoned person detection device 2. The abandoned person detection device 2 includes an imaging unit 20, a storage unit 21, an output unit 23, and a control unit 22.

  In FIG. 1, the abandoned person detection device 2 is provided on the ceiling of the monitoring area, but only the imaging unit 20 may be provided on the ceiling, and other devices may be provided in another place.

(Imaging unit 20)
The imaging unit 20 is a monitoring camera using, for example, a CCD imaging device. The imaging unit 20 is usually attached to the ceiling of the monitoring space (monitoring area). The shooting may be continuous shooting such as video shooting, but it is preferable that the monitoring area is sequentially shot at predetermined time intervals. The captured monitoring images of the monitoring space are sequentially output to the control unit 22. The time interval at which the monitoring image is captured is 1/5 second, for example. Hereinafter, the unit of time recorded at the time interval of imaging is referred to as time. In the present embodiment, the imaging unit 20 includes a fisheye lens and is installed with its optical axis directed vertically downward, and images the entire room as a monitoring space.

(Storage unit 21)
The storage unit 21 is a storage device such as a ROM (Read Only Memory) or a RAM (Random Access Memory). A flash memory or a hard disk may be used. The storage unit 21 stores various programs and various data, and inputs / outputs such information to / from the control unit 22. The various data stored in the storage unit 21 includes a background image 210 necessary for background difference processing and a posture-specific person model 211 necessary for posture determination of the tracking person. The captured image may be stored in a large capacity storage device such as a hard disk as it is.

  The posture-specific person model 211 is obtained by modeling the shape (outer shape) of a person in each posture by roughly dividing the posture of the person into two positions of standing / inverting. The posture-specific person model 211 includes a standing model 211R and an inverted model 211T.

  The standing model 211R and the inverted model 211T are shown in FIGS. For each model, the control unit 22 extracts the person area from the monitoring image obtained by the imaging unit 20, and applies the posture-specific person model 211 to the extracted person area, thereby changing the posture of the person being imaged. Used for estimation.

The standing model 211R is a posture-specific person model 211 simulating the shape of a person in a standing posture. In the present embodiment, the standing model 211R is defined by an ellipse having a major axis length L _R ± α _LR and a minor axis length S _R ± α _SR , as shown in FIG. In the coordinate system of the monitoring image captured by the imaging unit 20 with the optical axis directed vertically downward, the radiation direction from the center of the monitoring image is the height direction. ± αθ _R is a value for considering the inclination of the body, and the standing model 211R is arranged by limiting the major axis direction within the range of ± αθ _R from the radiation direction. That is, a person region in which the deviation (deviation angle) in the major axis direction from the image center in the radial direction is within ± αθ _R is applied.

Reference value L _R of the long axis length is set according to the average height, the variation allowable range ± alpha _LR is set to absorb possible range errors like caused by individual differences in physique. The short axis length reference value S _R is set according to the average body width, and the variation allowable range ± α _SR is within a range that can absorb errors caused by the inclination of the body and errors caused by individual differences in physique. Is set. The variation tolerance range ± αθ _R of the deviation angle of the long axis is set to a range in which an error caused by the body inclination and an individual difference in the standing posture can be absorbed. These permissible fluctuation ranges mean the shape and inclination range of the range that can be regarded as standing.

In the storage unit 21, the long axis length L _R , the short axis length S _R , the long axis length fluctuation tolerance ± α _LR , and the short axis length fluctuation tolerance ± preset as the standing model 211R as described above. Each numerical value of α _SR and the long axis deviation angle variation allowable range ± αθ _R is stored. Since the size of the person image changes depending on the positional relationship between the imaging unit 20 and the person, L _R and S _R are set for each position on the monitoring image, and α _LR and αS _R are ratios to L _R and S _R. It is better to set as

The inverted model 211T is a posture-specific person model 211 that imitates the shape of a person in a falling posture. In the present embodiment, the inverted model 211T is defined by an ellipse having a major axis length LT ± αLT and a minor axis length ST ± αST, as shown in FIG. ± αθ _T is a value for considering the inclination of the body. In the coordinate system of the monitoring image captured by the imaging unit 20 with the optical axis directed vertically downward, the inversion model 211T has its long axis direction in the radiation method. Arrangement is limited to a range of ± αθ _T from the line direction.

The major axis length reference value LT is set according to the average height, and the variation allowable range ± αLT is set to a range that can absorb errors caused by individual differences in physique. The short axis length reference value ST is set according to the average body width, and its variation tolerance range ± αST is set to a range that can absorb errors caused by body inclination and errors caused by individual differences in physique. The The variation allowable range ± αθ _T of the major axis deviation angle is set in a range of 0 ≦ αθ _T <90 ° −αθ _R so as not to be confused with the standing position. These permissible fluctuation ranges mean the shape and inclination range of the range that can be regarded as inversion.

In the storage unit 21, the long axis length LT, the short axis length ST, the long axis length variation allowable range ± αLT, and the short axis length variation allowable range ± αST, which are set in advance as the inverted model 211T as described above. Each numerical value of the variation allowable range ± αθ _T of the shaft deflection angle is stored. Since the size of the person image varies depending on the positional relationship between the imaging unit 20 and the person, LT and ST are set for each position on the monitoring image, and αLT and αST are preferably set as a ratio to LT and ST. .

(Control unit 22)
The control unit 22 is configured using an arithmetic device such as a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or an MCU (Micro Control Unit), and reads out a program from the storage unit 21 and executes it to extract a person area. Means 220, person tracking means 221, posture determination means 222 that functions as an inverted person determination means, stationary state determination means 223, slow speed movement determination means 224, and abnormality determination means 225 that functions as a person leaving state detection means.

  Here, each means of the control part 22 is demonstrated with the processing flow shown in FIG.

(Output part)
The output unit 23 is a communication unit that outputs an abnormal signal to the outside of the abandoned person detection device. When an abnormality signal is input from the abnormality determination unit 225 of the control unit 22, the output unit 23 outputs the abnormality signal to the controller 3.

“Processing Flow in Abandoned Person Detection Device 2”
FIG. 5 is a flowchart showing an abnormality detection process in the abandoned person detection device 2. Various data processing is basically performed in the control unit 22.

(S1: Initialization)
When the power is turned on, each unit is initialized and starts operating. The initialization includes a process of storing the monitoring image immediately after startup in the storage unit 21 as the background image 210. Here, it is preferable to acquire the background image every time when a person does not enter, but various means such as storing a fixed screen to some extent can be employed.

(S2: Monitor image acquisition)
When the initialization is completed, a monitoring image is acquired. That is, the imaging unit 20 outputs a monitoring image every time the monitoring space is imaged, and the control unit 22 acquires the monitoring image. In S2, a monitoring image at the current time is acquired and sent to the control unit.

(S3: Background / Labeling)
The person area extraction unit 220 of the control unit 22 removes the background from the monitoring image at the current time sent from the imaging unit, extracts the person area, and labels the obtained person area. The person area extraction unit 220 compares the monitoring image obtained from the imaging unit 20 with the background image 210 stored in the storage unit 21, and extracts a pixel whose difference from the background image 210 is greater than or equal to a threshold value as a change pixel. To do. Note that brightness, color, edge strength / direction, and the like can be used as the difference feature amount. As the threshold value, a predetermined value determined in advance may be used, or may be changed dynamically.

  Further, it is desirable to remove light / shadow regions by normalized correlation after the difference calculation, and to remove sesame salt-like noise by expansion / contraction processing.

  The person area extraction unit 220 performs labeling on the set of changed pixels after extracting the changed pixels. Labeling is a process of assigning the same number to pixels that are spatially connected, and an individual number is assigned to each block (set) of changed pixels. After the labeling, noise removal processing is performed again, such as removing one labeled region, for example, a block whose area is equal to or less than the threshold value. The threshold value at this time is determined from the standard person size calculated from the camera installation conditions. Hereinafter, this block is referred to as a person area.

(S4: Person tracking)
The image of the person area obtained by the person area extracting unit 220 is sent to the person tracking unit 221. The person tracking means 221 is extracted by the person area extracting means 220 by the previous time (in the monitoring image acquired before the previous time) and stored in the storage unit 21 with the tracking characteristics and the current time ( The person area is tracked by associating it with each person area extracted by the person area extracting means 220 (in the monitoring image acquired this time).

  Therefore, whenever the appearance of the person area in the monitoring area is confirmed, the person tracking unit 221 extracts the image feature of the person area as a tracking feature and stores it in the storage unit 21. Then, the tracking feature of the person region extracted at the previous time stored in the storage unit 21 is compared with the tracking feature of the person region extracted at the current time, and the person region determined to be the most similar Associate.

  The tracking feature includes a color and edge histogram, a gravity center position of a person area, a circumscribed rectangle, and the like. If the person area extracted at the current time is not associated with any of the tracking features stored in the storage unit, it is determined that the person image is a person area newly appearing at the current time. .

  When there is a person area that is not associated with the person area extracted at the current time among the tracking features stored in the storage unit 21, the tracking information has moved out of the field of view of the input image. Is determined to belong to a person who has left the monitoring area such as an office, and is deleted from the storage unit 21. In addition, it is good also as what keeps fixed time and tries said correlation process in the meantime.

  The process in the person tracking unit 221 described above is a general tracking process in the field of image processing technology, and thus the details are omitted.

(S5-S6: Posture determination)
The posture determination unit 222 selects a posture-specific person model 211 that best matches the shape of the person region extracted by the person region extraction unit 220, and the posture corresponding to the selected posture-specific person model 211 is included in the person region. Judged as the posture of a person. When a plurality of person regions are extracted, the posture determination unit 222 performs posture determination for each person region (S5). The posture determination unit 222 functions as an inverted person detection unit that detects a person in an inverted posture.

The posture determination unit 222 calculates the fitness by superimposing the posture-specific human models 211 on the person area, and selects the posture-specific human model 211 having the highest fitness. In this case, the fitness is expressed by the following evaluation function.
(Evaluation function) =-{(Pixel that is a person area but not a person model area)
+ Γ × (pixel model region but not human region)}

  The first term is the number of pixels that are not covered by the person model despite the person area, and the second term is the number of pixels that are not covered by the person model but is not the person area. Is evaluated by the number of protruding pixels. The evaluation function takes a negative value, and the higher the value (the closer to 0), the higher the matching degree between the person area and the person model.

  γ is a parameter for balancing the first term and the second term. For example, when a threshold value for difference processing is set so that a changed pixel is easily extracted, the value of γ may be increased.

  Here, the posture of each person is not constant, and the positional relationship between the persons is not constant. As described above, the posture of the person includes errors caused by individual differences and body inclinations. Therefore, the process of superimposing the individual person models 211 for each posture on the person region is performed in a search manner by repeating the process while changing each parameter. However, in order to detect anomalies in real time, the number of iterations N is fixed in advance so that the processing is completed at one time in advance, or the number of iterations N is dynamically determined in advance by defining a processing time shorter than one time. Stipulated in For example, the number of repetitions N is defined as 1000. At this time, in order to prevent the result of posture determination from falling into a local solution within a limited number of iterations, it is preferable to change the parameters at random.

  Next, parameters that are changed during the search will be described. The posture determination means 222 tries a plurality of posture-specific person models 211 corresponding to each posture in response to the posture that each person takes is not constant. In the present embodiment in which two types of postures (standing position and inverted position) are set, a random selection is made from two combinations.

  In addition, the posture determination unit 222 randomly changes the shape and inclination of the posture-specific person model 211 in each posture within a variation allowable range in response to errors caused by individual differences and body inclinations. The shape is changed by independently changing the major axis length and the minor axis length. The inclination is changed by changing the deflection angle.

Specifically, the posture determining section 222, a gravity center position P _R of the standing model 211R sets the random coordinates the person area, the long axis length corresponding to the center-of-gravity position P _R L _R and minor axis length S _R Are read from the storage unit 21 and the allowable ranges α _LR , α _SR , αθ _R are read, Δ _LR is randomly set in a range of ± α _LR with respect to the long axis length L _R , and the short axis length S _R is set. set the delta _SR in the range of ± alpha _SR randomly set the [delta] [theta] _R randomly in a range of ± αθ _R to the long axis deflection angle. Posture determination unit 222, a line rotated from the radiation connecting the image center 83 and the position of the center of gravity P _R around the center-of-gravity position P _R by [Delta] [theta] _R determined as the major axis, the major axis length to the center of gravity position P _R L _R + delta _LR, and calculates the ellipse minor axis S _R + Δ _SR.

Similarly, the posture determination means 222 randomly sets the center of gravity position P _T of the inverted model 211T in the person area, and sets the fluctuation amounts Δ _LT , Δ _ST , Δθ _T to ± α _LT , ± α _ST , ± αθ, respectively. _R is randomly set within the range of _T, a normal passing through the gravity center P _{K of the} radiation connecting the image center 83 and the gravity center P _T is obtained, and a straight line rotated by Δθ _T around the gravity center P _T from the normal is obtained. An ellipse having a major axis length L _T + Δ _LT and a minor axis length S _T + Δ _ST centered on the center of gravity position P _T is calculated.

  Then, the posture determination unit 222 calculates the number of pixels in a portion that does not match between the person area and the person model area, and calculates the fitness using the evaluation function. With this operation, if the person model having the highest fitness in the person area is a person model in the inverted posture, the posture of the person included in the person region is determined to be the inverted posture (S6).

  In this way, the posture determination unit 222 obtains the posture of the person area as the tracking result as the tracking feature, and the obtained tracking feature is stored in the storage unit 21.

  Also, continuity in the time direction can be introduced into the above evaluation function that represents the goodness of fit between the person region and the person model. For example, the color histogram in the person model area at the previous time is calculated, and the degree of coincidence with the color histogram in the person model area this time is added to the evaluation function so that the color does not vary greatly between frames. An ellipse can be fitted.

  When a plurality of persons are included in one extracted person area, it is only necessary to fit as many person models as the number of persons to the person area while changing each parameter.

  When only one person is included in one person area, it can be determined that the person area is in the inverted posture without using the person model for each posture. In that case, the human body region is approximated by an ellipse or the like, and the inverted posture is determined from the inclination of the long axis. How a person in an inverted position is projected on the screen depends on the camera used and its installation conditions.For example, with a fish-eye camera installed vertically downward on the ceiling surface, the straight line connecting the center of the image and the center of the person area If the major axis of the approximate ellipse is almost orthogonal, it is determined as the inverted posture. At this time, an allowable range may be provided for the angle, and the inverted posture may be determined when the angle formed by the long axis and the radiation connected to the center is 90 ° ± δθ. For example, δθ is 15 °.

  As described above, the person in the inverted posture can be detected in S5 to S6.

(S7-S8: Stillness determination)
In S6, when an inverted person is detected, the stationary state determination unit 223 scans back the tracking result of the person area, and determines whether or not the person is still standing in the inverted posture for a predetermined frame or more. Is determined (S7). The following criteria are used to determine whether a person is still.
(I) The center-of-gravity position coordinates of the person area have not changed.
(Ii) The size of the circumscribed rectangle of the person area has not changed.
As the simplest determination method, a method of analyzing that the position of the center of gravity and the size of the circumscribed rectangle are unchanged from the frame one hour before may be used for the human region in the inverted posture.

  It is desirable to set an appropriate threshold for the change amount for both (i) and (ii) so as not to be sensitive to minute changes due to differential noise or the like.

When the conditions (i) and (ii) are satisfied, the stationary counter is counted up as being stationary (S8). If the condition is not satisfied, the process shifts to the slow movement determination in S9 because it is not stationary.

  At this time, it may be determined whether the condition of (i) and (ii) is satisfied after confirming that the frame one hour before is also in the inverted posture. In order to realize more stable detection, it is desirable to observe stillness over several frames. The tracking result may be traced back in the span where the inverted posture continues, and the stillness in that period may be determined using the conditions (i) and (ii).

(S9 to S10: Slow movement determination)
In S7, when the person in the inverted posture is not stationary, the slow movement determination means 224 calculates the speed from the center of gravity position coordinates at the previous time and the center of gravity position coordinates at the current time, which is a predetermined threshold value. If it is the following, it is determined that it is moving at a slow speed (S9). Although the threshold varies depending on the camera to be used and installation conditions, in the case of a fish-eye camera installed on a ceiling surface of 2.5 m, it is desirable to set it to about several pixels around the fish-eye and about a dozen pixels around the center. In addition, assuming that the straight line connecting the center of the fisheye image and the center of gravity of the person area and the person area (the intersection closer to the center) is the person ground plane, the 3D coordinates are calculated using the camera installation conditions. The slow speed may be determined based on the change in the three-dimensional coordinates.

  If it is determined in S9 that the movement is slow, the fine speed counter is incremented (S10). If it is not a slow movement in S8, the counter is reset in S12 because it is not stationary and is not a slow movement. It should be noted that the counter is also reset in S12 when a person in an inverted posture is not detected in S6. Here, the counters to be reset are all of a stationary counter, a fine speed counter, and a neglect counter.

  As another criterion for determining the slow speed movement in S9, the average speed may be obtained by dividing the total moving distance from the time of the inverted posture by the number of observation frames, and the slow speed determination may be performed based on the average speed.

  In the above description, only one person in the inverted posture is the target, but when there are a plurality of persons, stationary and slow movement are determined for each.

(S13: Abandoned person determination)
In this manner, the stationary counter and the fine speed counter are counted up in S8 and S10. Then, the neglected counter adds the weighted count values of the stationary counter and the count value of the fine speed counter. That is, the standing counter is counted up with a greater weight on the count value of the stationary counter (S11).

  As a result, the number of frames that are stationary in the inverted posture or moved at a slow speed in the inverted posture is weighted and added to the neglected counter. The abnormality determination unit 225 determines that there is an abandoned person when the value of the abandonment counter is equal to or greater than a predetermined threshold (S13). For example, as the weight, the slow movement is set to 1, the stillness is set to 1.2, and the threshold value of the neglect counter is set to 120 seconds.

  As another device for counting up, instead of always adding a constant value, the degree of stillness and the degree of slow movement can be expressed as continuous values, and the number to be counted up can be made variable according to the values. For example, the difference from the predetermined threshold can be counted up with respect to the slow moving speed.

  Furthermore, it is also possible to make a determination by individually setting threshold values for the stationary counter and the slow movement counter. For example, when the count value of the stationary counter reaches 90 seconds or the count value of the slow movement counter reaches 120 seconds as a result of the individual count, it may be determined that the abandoned person is detected.

  Further, as another embodiment, it is possible to realize a mechanism for outputting an abnormal signal when the moving distance due to the slow movement in the inverted posture exceeds a predetermined threshold.

  As described above, the abnormality determination unit 225 functions as a person leaving state detection unit, and determines that an abnormality is detected when a person in the state of being left (a left person) is detected.

  Furthermore, as another embodiment in consideration of an error in posture determination, as a counter reset in S12, the counter is not reset at a stroke when the inverted posture is lost, but only a predetermined value is left when the posture is not inverted. The count value may be subtracted. For example, the counter addition value when the stillness in the inverted posture is observed is set to 2, and 1 is subtracted when the posture is not the inverted posture. The subtraction value in the neglected counter is desirably subtracted by making the subtraction value for the count value of the stationary counter different from the subtraction value from the count value of the fine speed counter, but the subtraction values of both may be the same value.

(S14: Abnormal output)
When it is determined in S13 that there is an abandoned person, the abnormality determination unit 225 outputs an abnormality signal to the output unit 23.

"Effect of this embodiment"
As described above, according to the abnormality detection device according to the present embodiment, it is detected that the person area is in a collapsed posture, and the posture of the person area remains in the collapsed posture even when the person region is still in the collapsed posture. Even when moving at a slow speed, it is counted up as a neglected state, and when the counter exceeds a predetermined threshold, it can be determined that the person is a neglected person.

  DESCRIPTION OF SYMBOLS 1 Report system, 2 Abandoned person detection apparatus, 3 Controller, 4 Wide-area communication network, 5 Center apparatus, 6 Safe, 20 Imaging part, 21 Storage part, 22 Control part, 23 Output part, 83 Image center, 210 Background image, 211 Posture-specific person model, 220 person region extraction means, 221 person tracking means, 222 posture determination means, 223 stationary state determination means, 224 slow movement determination means, 225 abnormality determination means.

Claims (3)

An image acquisition unit that sequentially acquires monitoring images obtained by photographing the monitoring area, a storage unit that stores a monitoring image in a state where no person is present in the monitoring area as a background image, and a processing of the monitoring image acquired by the image acquisition unit An abandoned person detection device configured from a control unit that performs
The controller is
Person area extracting means for extracting a person area from the monitoring image and the background image;
An inverted person detecting means for detecting a person area in an inverted state that is a posture in which the person has fallen out of the person area;
A stationary state determination unit that temporally tracks the human region in the inverted position and determines whether the human region is in a stationary state that is not substantially moved;
Wherein temporally tracking the person area of倒位state, it is determined that the person area is not in the stationary state, and a very low speed movement determining means determines that a very low speed moving state when moving predetermined speed following a very slow speed,
Based on the determination results of both the stationary state determination unit and the slow speed movement determination unit, a person leaving state detection unit that detects a state in which the person is left according to a predetermined rule ;
An abandoned person detection device characterized by comprising: An image acquisition unit that sequentially acquires a monitoring image obtained by photographing the monitoring area via a fisheye lens from above the monitoring area, and a monitoring image in a state where no person is present in the monitoring area is stored as a background image , and a major axis and a minor axis are stored. A neglected person detection device comprising: a storage unit that stores a long and narrow figure as a person model corresponding to an outline of a person; and a control unit that processes a monitoring image acquired by the image acquisition unit,
The controller is
Person area extracting means for extracting a person area from the monitoring image and the background image;
Comparing the person area with the person model, a person area in an inverted state is a posture that has fallen when the major axis of the person area coincides with a person model oriented in a direction orthogonal to the radial direction from the center of the image. An inverted person detecting means for detecting;
A stationary state determination unit that temporally tracks the human region in the inverted position and determines whether the human region is in a stationary state that is not substantially moved;
A slow movement determination means that temporally tracks the person area in the inverted position and determines that the person area is moving at a slow speed of a predetermined speed or less;
Based on the determination results of both the stationary state determination unit and the slow movement determination unit, a person leaving state detection unit that detects a state in which the person is left;
An abandoned person detection device characterized by comprising: In the abandoned person detection device according to claim 1 or 2 ,
The person leaving state detecting means is
When the elapsed time in the stationary state exceeds a first predetermined value or when the elapsed time in the slow movement state exceeds a second predetermined value that is longer than the first predetermined value, it is detected that the person is left alone. An abandoned person detection device characterized by the above.

Images