JPH1169342A - Image processing system for tracking intruding object - Google Patents

Abstract

PROBLEM TO BE SOLVED: To control a photographing lens that can zoom to magnify an intruding object and to enable a monitoring person to easily confirm the intruding object on the screen of a video monitor by tracking and photographing automatically an object entering a required monitoring part. SOLUTION: An image processor 3 converts the position coordinates of an intruding object detected from an image signal having a wide visual field into a moved variable of a motor-driven pan head 18 which can turn in both vertical and horizontal directions. Thus, the processor 3 controls the head 18. A TV camera 17 containing a photographing lens 16 and put on the head 18 that is controlled toward the position of the intruding object outputs the image signals of the magnified intruding object to a video monitor 19. Thus, the magnified introducing object can always be monitored by the monitor 19. Therefore, a monitoring person can accurately confirm the expression, shape, etc., of the intruding object which are hitherto difficult to decide in an image of a wide visual field that is monitored by a video monitor 13.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for automatically detecting an intruding object from an image signal picked up and outputted by a television camera, tracking the intruding object and enlarging the invading object on a video monitor screen, thereby facilitating monitoring of the intruding object. The present invention relates to an intruding object tracking image processing system.

[0002]

2. Description of the Related Art The configuration and operation of a system for automatically detecting and tracking an intruding object using a television camera according to the prior art will be described with reference to FIGS. FIG.
FIG. 7 is a system diagram illustrating a system for automatic detection and tracking for detecting an intruding object according to the related art, and FIG. 7 is a diagram illustrating processing timing in a conventional system for automatic detection and tracking of an intruding object. The system for automatically detecting and tracking an intruding object includes a television camera 602 having an imaging lens 601, an image processing device 603, and a video monitor 614.

The structure of the image processing device 603 of the automatic detection and tracking system shown in FIG.
An A / D converter 4 converts an image signal input from the television camera 602 into a digital image signal.
Reference numeral 05 denotes a memory control unit that switches the digital image signal input from the A / D conversion unit 604 at a required portion and outputs the digital image signal to two image memories alternately.
Reference numeral 61 denotes an image memory for recording a digital image signal switched and input by the memory control unit 605;
Reference numeral 5 denotes a calculation unit that calculates the image signals read from the two image memories 606 and 607 and outputs the difference, and 608.
Denotes an image memory for storing the difference between the image signals obtained by the arithmetic unit 615; 616, a binary conversion unit for binarizing the difference image signal read from the image memory 608;
Is an image memory for storing the binarized signal binarized by the binarization conversion unit 616, and 612 is a memory control unit 6
A CPU 611 for switching the image signal to the two image memories in 05 and determining the presence / absence of an intruding object based on the binarized signal; 611, a character generating unit for generating a required character based on data input from the CPU 612; 610
Is an image signal output unit that combines a digital image signal input from the A / D conversion unit 604 with a character signal input from the character generation unit 611, converts the image signal into an analog image signal, and outputs the analog image signal. 1 shows an I / O control unit that outputs an intruding object detection signal based on a judgment signal indicating that an intruding object is present.

[0004] The television camera 602 captures an image of a required monitoring location via the imaging lens 601 and outputs an image signal to the image processing device 603. When an intruding object enters the monitoring location, the invading object is captured by the television camera 602 via the imaging lens 601,
An image signal of the monitoring location including the intruding object is output from the television camera 602 to the image processing device 603. The image signal output from the television camera 602 to the image processing device 603 is an image signal having a frame period of 1/30 second shown in 701 in FIG.

[0005] The image signal input to the image processing device 603 is first input to the A / D converter 604, where it is converted into a digital image signal.
5 and the image signal output unit 610. The memory control unit 605 switches the input digital image signal at a required portion according to a selection signal separately input from the CPU 612, and outputs the signal to the image memory 606 and the image memory 607 alternately. The image memory 606 and the image memory 607 store image signals of a required portion which are switched and alternately input by the memory control unit 605, respectively. The timing at which the image signal input to the memory control unit 605 is switched by the selection signal and alternately input to the image memory 606 and the image memory 607 and stored therein is determined by the image memory 60.
7 and 702, and the timings stored in the image memory 607 are shown in FIGS. 7 and 703. The image signal 708 is stored in the image memory 606, and the image signal 709 is stored in the image memory 607. Is done.

The image signals stored in the image memory 606 and the image memory 607 are read out at the same timing and output to the arithmetic unit 615. Arithmetic unit 6
Numeral 15 performs an operation on the input image signal from the image memory 606 and the image signal from the image memory 607, and
The difference between the two image signals is calculated and output to the image memory 608. The calculation by the calculation unit 615 is performed by the image signal 708 shown in FIGS. 7 and 702 and the image signal 709 shown in FIGS.
The difference 710 between the two image signals shown in FIGS. 7 and 704 is output to the image memory 608. The next operation is performed by using the image signal 709 shown in FIGS.
7, and the difference 712 between the two image signals 704 shown in FIGS. If there is no intruding object at the required monitoring location being imaged and there is no change, of course, the difference between the two image signals is 7
10, 712... Do not occur.

[0007] The image memory 608 stores an image signal of the difference between the two input image signals, and outputs the read image signal to the binarization conversion unit 616. Binary conversion unit 616
Converts the difference between the two image signals read from the image memory 608 into a binary image signal and outputs the binarized image signal to the image memory 609. The binarized image signal stored in the image memory 609 is a binary image signal shown in FIGS.
.. Are shown as coded image signals 713, 714. The time difference between the signals in FIGS. 7 and 704 and the signal in FIGS. 7 and 705 is a time difference generated in the image memory in which the signals are stored.

The image memory 609 stores the binarized image signal and outputs the read binarized image signal to the CPU 612. The CPU 612 calculates the size of the intruding object based on the input binary image signal, determines whether or not there is an intruding object, outputs data according to the determination result to the character generating unit 611, and when there is an intruding object, Then, an intrusion object detection signal, for example, an alarm signal is output to an external device via the I / O control unit 613. The character generation unit 611 is
A character signal is generated based on the data output from 12, and output to the image signal output unit 610. Image signal output unit 610
Synthesizes an input digital image signal and a character signal, converts them into an analog image signal, and outputs the analog image signal to another device, for example, a video monitor 614.

A conventional method for tracking an intruding object detected in this manner will be described using an intruding object detection system shown in FIG. This intruding object detection system includes a plurality of television cameras 801 and 8 for detecting an intruding object.
02, 803, 804 and a plurality of television cameras 8
01, 802, 803, 804 An image signal switch 805 for switching each image signal output from the image signal.
And an image processing device 806 for processing an image signal switched by the image signal switch 805 to detect an intruding object.
And a video monitor 807 for the observer to check the intruding object that has entered the required monitoring location.

[0010] A plurality of television cameras 801, 80
Reference numerals 2, 803, and 804 are provided so as to capture images of required monitoring points and to monitor the intrusion of an intruding object as a continuation of the field of view of each television camera, and the television cameras 801, 802, 803, and 804 The captured image signal is output to the image signal switch 805. For example, when an image processing start signal 809 is input by an operation from a monitor, the image processing apparatus 806 monitors an intruding object at a required monitoring location, for example, a television camera 801.
A switching control signal 810 is output to the image signal switch 805 in order to monitor the image signal on the screen of the video monitor 807. The image signal switch 805 outputs the switching control signal 81
The image signal of the television camera 801 switched by 0 is output to the image processing device 806. The image processing device 806 performs image processing based on the input image signal of the television camera 801 to detect an intruding object, and outputs an image signal of the intruding object to the video monitor 807. When the intruding object goes out of the field of view of the television camera 801, the image signal of the next television camera 802 is switched to continue the monitoring.

[0011]

In an intruding object tracking image processing system according to the prior art, when tracking an intruding object, a plurality of television cameras are used, and respective images output from the plurality of television cameras are used. Although the detection of an intruding object is performed while switching signals, the number of television cameras increases, and furthermore, a controller for switching with an image signal switch is required, thus increasing the scale of the system. There is. In order to reduce the number of television cameras in order to reduce the size of the system, it is necessary to widen the field of view taken by one imaging lens and the television camera. For example, when the intruding object is a human, it is difficult to determine the human phase of the intruding human. The present invention solves the above-mentioned problems, performs automatic detection of an intruding object, and controls a pan head rotating in the vertical and horizontal directions based on the detected and calculated data to automatically track the intruding object and perform imaging capable of zooming. It is an object of the present invention to provide an intruding object tracking image processing system which controls a lens to enlarge an intruding object and makes it easy for a monitor to check the intruding object on a video monitor screen.

[0012]

In order to achieve the above object, an intruding object tracking image processing system according to the present invention comprises a television having a wide-field imaging lens for outputting an image signal used for detecting an intruding object. A television camera equipped with a camera, a zoomable imaging lens that is mounted on the electric pan head, tracks an intruding object, and outputs an image signal used for image monitoring, and a television equipped with the wide-field imaging lens An image processing device that processes and calculates an image signal input from a camera and outputs a rotation control signal for rotating the electric pan and head by a required amount, and detects an intruding object that has entered a required monitoring location. , And automatically captures an image.

Further, the intruding object tracking image processing system according to the present invention comprises a television camera having a wide-field imaging lens for outputting an image signal used for detecting an intruding object;
A television camera equipped with a zoomable imaging lens that is mounted on an electric pan head and tracks an intruding object and outputs an image signal used for enlarged image monitoring, and a television camera equipped with the wide-field imaging lens An image which processes and calculates an image signal input from the CPU and outputs a rotation control signal for rotating the electric pan head by a required amount and a zooming control signal for zooming the zoomable imaging lens by a required amount. A processing device for automatically tracking and automatically zooming in on an intruding object that has entered a required monitoring location.

More specifically, the image processing apparatus of the intruding object tracking image processing system of the present invention stores an image signal from a television camera having an imaging lens with a wide field of view and an image signal obtained by processing the image signal. An image memory, a CPU for performing processing, calculation, control, management, and the like in the image processing apparatus; a program memory for storing a program for operating the CPU; A work memory for storing data for analyzing the data obtained, and a rotation control signal for rotating the electric pan head by a required amount based on the calculated data and a required amount of zooming of the zoomable imaging lens. And a pan head controller for outputting the zooming control signal.

More specifically, the image memory of the image processing apparatus of the intruding object tracking image processing system according to the present invention comprises:
An input image memory for storing an input image signal, a background image memory for storing a background image signal used for detecting an intruding object, and a background image signal and an input image signal for performing an operation for detecting an image difference. Two work memories for storing the respective images, an inter-image operation result image memory for storing a result obtained by detecting and calculating the difference between the background image and the input image, and a binary obtained by detecting and calculating the difference between the images. And a binarized image memory for storing the digitized image signal.

An intruding object tracking image processing system of the present invention processes image information input from a television camera having an imaging lens with a wide field of view to an image processing apparatus, detects an intruding object, and detects the detected intruding object. Based on the position information and size information of the camera, the electric pan head equipped with a television camera equipped with a zoomable imaging lens is controlled to rotate in the direction of the intruding object, and the zooming of the zoomable imaging lens is controlled. This makes it easier for the observer who monitors the video monitor to determine the intruding object.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an intruding object tracking image processing system according to the present invention will be described with reference to FIGS. FIG. 1 is a system diagram showing an intruding object tracking image processing system according to the present invention, and FIG. 2 is an explanatory diagram of images taken by an imaging lens and a television camera having different imaging fields of view, which constitute the system. An intruding object tracking image processing system according to the present invention includes a television camera 2 having an imaging lens 1 for obtaining an image of a wide field of view.
A television camera 17 equipped with a zoomable imaging lens 16 mounted on an electric pan head 18 for obtaining an enlarged image; and detecting an intruding object by performing image processing on an input image signal of a wide field of view. An image processing device 3 for performing necessary control to obtain an enlarged image, a video monitor 13 for monitoring images having different imaging fields of view,
19.

In the intruding object tracking image processing system shown in FIG. 1, an imaging lens 1 for obtaining an image of a wide field of view.
The television camera 2 provided with the image processing unit converts an image signal obtained by capturing an image of a required monitoring location and having a wide field of view as shown in FIG. 2A for image signal processing for detecting an intruding object. Output to the device 3. In addition, the television camera 1 having the imaging lens 16 capable of zooming.
Reference numeral 7 is mounted on an electric pan / tilt head 18 that can rotate in the vertical and horizontal directions. The image processing device 3 controls the required amount of vertical and horizontal directions and zooming, and controls the required monitoring location. FIG. 2B shows an image of an intruding object, for example.
The image signal of the enlarged image as shown in FIG. The image processing device 3 detects an intruding object from an image signal of a wide field of view input from the television camera 2 including the imaging lens 1 and calculates the position coordinates of the intruding object on the captured image. further,
The image processing device 3 controls the electric pan head 18 by converting the position coordinates of the intruding object detected from the image signal having a wide field of view into the movement amount of the electric pan head 18 that rotates vertically and horizontally. . Electric head 1 controlled in the direction of the intruding object
A television camera 17 having an imaging lens 16 mounted on the camera 8 to output an image signal of an enlarged image of the intruding object to a video monitor 19, and the video monitor 19 constantly monitors the enlarged intruding object. Can be. By monitoring the intruded object enlarged by the video monitor 19, the observer clearly confirms the expression, shape, and the like of the intruded object, which were difficult to determine based on the image of the wide field of view monitored by the video monitor 13. be able to.

Next, a circuit constituting the image processing apparatus 3 and its operation will be described. An image signal output from a television camera 2 having an imaging lens 1 for obtaining an image of a wide field of view is input to an image processing device 3. The image signal input to the image processing device 3 is first input to an image signal input interface (I / F) 4, and is converted from an analog signal to a digital signal of 256 tones from 0 to 255, for example, and the data bus 12 Output to An image memory 5 is connected to the data bus 12. The image memory 5 includes an input image memory 51 for recording an input image signal, a background image memory 52 for recording a background image signal used for detecting an intruding object, and an image memory. Work image memories 53 and 54 for respectively recording a background image signal and an input image signal for performing an operation for detecting a difference, and performing a required operation between image memories, and a required operation for detecting a difference between images. An image calculation result image memory 55 for recording the result,
It has a binarized image memory 56 for storing a binarized image signal obtained by binarizing a result of a required operation, and a capacity capable of storing six image memories.

The data bus 12 has arithmetic, control,
A CPU 6 for management and the like; a program memory 10 storing a program for operating the CPU 6; a work memory 9 for analyzing data obtained by image signal processing such as position coordinates of detected intruding objects, action history; The CPU 6 calculates the amount of movement of the electric camera platform from the position coordinates of the intruding object, and outputs a rotation control signal for rotating the motorized camera platform by a required amount under the control of the CPU 6 and the required size according to the size of the intruding object. A motorized pan head controller 8 for outputting a zooming control signal for controlling the required amount of zooming of the imaging lens 16 so as to obtain an intruding object image, an alarm signal when an intruding object is detected, a measurement start signal, a control signal, etc. An external I / F 11 that interfaces with external devices for input and output, converts digital image signals to analog image signals, and Video output I to be output to the Omonita, etc.
/ F7 is connected.

FIG. 4 is a flowchart showing a processing procedure for an embodiment of an intruding object tracking image processing system according to the present invention, and the processing procedure will be described. In the background image creation processing 401, the image signal having a wide field of view without the intrusion of the intruding object at the required monitoring point output from the television camera 2 is stored in the background image memory 52 at predetermined frame intervals,
A weighted average calculation of the stored image signals is performed.
There is no intruding object as shown in (a), or
An image that can be ignored even if there is any object, that is, a background image 301 is created, and the background image signal is stored. In the image input processing 402, an image having a wide field of view of a required monitoring location output from the television camera 2, for example, an image signal of an input image 302 as shown in FIG. The data is input to and stored in the input image memory 51 via the memory. In the difference processing 403, the input image memory 5
1 is copied to the work image memory 53, the image signal stored in the background image memory 52 is copied to the work image memory 54, the image signals are read out from the work image memory 53 and the work image memory 54, and the operation is performed. As a result, a difference image signal of the difference image 303 as shown in FIG. 3C is obtained as a calculation result, and this difference image signal is input to the inter-image calculation result image memory 55 and stored.

In the binarization process 404, the difference image signal read from the inter-image operation result image memory 55 is binarized, and the binarized image signal of the binarized image 304 as shown in FIG. Then, the binarized image signal is input to the binarized image memory 56 and stored. In the labeling process 405,
From the binarized image signal read out from the binarized image memory 56, the presence or absence of a cluster of independent white pixels is detected, and the number of clusters of white pixels is further measured. In the object recognition process 406, as shown in FIG. 3D, the size, for example, the width 309 in the Y direction and the width 310 in the X direction are measured for each cluster of white pixels whose number has been measured in the labeling process 405. The intruding object to be detected is recognized by comparing with a preset range of the intruding object's outer shape data, and the image memory position and size data of the detected intruding object (for example, the width 309 in the Y direction, the X direction) Width 3
10) is stored in the work memory 9.

In the object presence / absence determination process 407, the presence / absence of an intruding object is determined from the data stored in the work memory 9.
If it is determined that the intruding object is to be detected in the object presence / absence determining process 407, in the coordinate calculating process 408, for the intruding object stored in the work memory 9, as shown in FIG. Y direction 305, X direction 3
06 is calculated, and the number of objects and the coordinates of the position of the center of gravity of each object are stored in the work memory 9. At this time, if there are a plurality of detected objects, in the tracking determination process 409, the coordinates of the object's body weight and center, the size of the objects stored in the work memory 9, and the work memory 9 up to the previous process are stored. From the coordinates of the object's body weight and center, the size of the object,
The coordinates of the object's body weight and center position and the size of the object in the current input image obtained from the moving speed of the object are collated, and the object stored up to the previous time is limited.

In the motorized head control process 410, the coordinates of the object's body weight and center position of the obtained input image are converted into absolute position data of the motorized head 18, and the difference from the position where the motorized head 18 is currently stationary is calculated. Is calculated, and a rotation control signal is output from the pan head controller 8 to the electric pan head 18 so as to rotate the electric pan head 18 in the required direction by this difference. Next, in the background image update process 411, the input image is weighted with respect to the created background image by using a weighted average calculation expression shown in the following equation (1), and an updated background image is created. Background updated image = background image × weighting factor + input image × (1−weighting factor) (1) The weighting factor shown in equation (1) is a change in the input image with respect to the background image obtained so far. The updating of the background image can be improved by reducing the weighting coefficient. In addition, as a background image updating method other than the background image updating process, a method of recognizing an image when no intruding object exists and updating the background image can also be realized. In the alarm output process 412, after detecting an intruding object, an alarm signal is output to an external device via the external I / F 11, and an alarm is given to a supervisor.

Further, the calculation of the position coordinates of the object of the detected image and the pan head control data used in the pan head control process 410 will be described with reference to FIG. In the case of an image in which a person 506 has invaded in the upper left direction of the screen, for example, as in a detected image 501 shown in FIG. 5A, which is captured by the television camera 2, the barycentric position coordinate Y direction of the person 506 at this time 503, X direction 502 is calculated and obtained. Next, the motor-driven pan head 18 equipped with the television camera 17 having the zoomable imaging lens 16 for obtaining an enlarged image is controlled, and the tracking image 50 shown in FIG.
8 is input, that is, the intruding person 506 is positioned so as to be enlarged and displayed at the center of the screen of the video monitor 19, and the electric pan head 18 at that time is positioned.
Read the position data in the vertical and horizontal directions. Similarly, the coordinates Y505 and X50 of the center of gravity of the person 507 at the lower right position on the screen of the detection image 501 shown in FIG.
4 is obtained, and the vertical and horizontal position data of the electric platform 18 at the position where the tracking image 509 shown in FIG. A head control coefficient is calculated from the obtained position data of the electric head 18.

The head control coefficient is given by the following equation (2) and (3)
It can be obtained from the formula. Left-right direction pan head control coefficient = (left-right position of pan head to obtain image 509-left-right position of pan head to obtain image 506) / (X-direction position coordinate 504 of image 507-X-direction position of image 506) Coordinates 502) (2) Vertical head control coefficient = (vertical position of head to obtain image 509-vertical position of head to obtain image 508) / (Y-direction position of image 507) Coordinates 505—Y-direction position coordinates 503 of image 506) (3) The electric head position of the detection position is calculated by integrating the results obtained by Expressions (2) and (3) with the detection position. It becomes possible and control for automatic tracking becomes possible.

In the object presence / absence determination process 407, the camera platform control process 41 is performed based on the calculated size of the intruding object.
At 0, a zooming control signal can be output from the pan head controller 8 to the zoomable imaging lens 16 to change the focal length in order to obtain a screen that is automatically tracked and automatically enlarged. The size of the intruding object can be obtained in a predetermined size. The focal length of the zoomable imaging lens 16 is controlled by a value calculated by the following equation (4). Control focal length = (total number of pixels of memory in width 309 in Y direction / detection width 309 on memory in detection object Y direction) × focal length in intruding object detection screen 201 × viewing factor × (4)

In equation (4), the calculation is performed using the Y direction, but the same result can be obtained by using the X direction. However, if the intruding object is a human, X
Since the size of the width 310 in the direction and the width 309 in the Y direction is large in the Y direction, it is preferable to calculate in the Y direction in the control for zooming up on the basis of the entire human. Further, depending on the moving speed of the intruding object, the intruding object may deviate from the following screen. In such a case, it is possible to cope with the situation by expanding the field of view by the field factor. Thereby, even when the intruding object moves from far away to near to the camera, it is possible to image the intruding object with a fixed size. In the description of the intruding object tracking image processing system described above, the zoomable imaging lens 1 is used.
6 is controlled by the zooming control signal output from the pan head control unit 8 to obtain an automatically enlarged screen. However, the zooming control signal manually controlled by the observer is transmitted to the external I / F 1.
The enlarged screen may be manually adjusted to a required size by inputting the data to the pan head control unit 8 via the control unit 1.

[0029]

According to the present invention, automatic detection of an intruding object and automatic tracking and zooming of an intruding object by controlling an electric pan head rotating in the vertical and horizontal directions based on detected and calculated data are possible. It is possible to provide an intruding object tracking image processing system that controls a simple imaging lens to enlarge an intruding object and makes it easy for a monitor to check the intruding object on a video monitor screen.

[Brief description of the drawings]

FIG. 1 is a system diagram of an intruding object tracking image processing system according to the present invention.

FIG. 2 is a video monitor screen for explaining the intruding object tracking image processing system according to the present invention.

FIG. 3 is a video monitor screen for explaining the intruding object tracking image processing system according to the present invention.

FIG. 4 is a flowchart of a processing procedure of the intruding object tracking image processing system according to the present invention.

FIG. 5 is a video monitor screen for explaining the intruding object tracking image processing system according to the present invention.

FIG. 6 is a system diagram of an intruding object tracking image processing system according to a conventional technique.

FIG. 7 is an explanatory diagram of processing timing for explaining an intruding object tracking image processing system according to a conventional technique.

FIG. 8 is a system diagram of an intruding object tracking image processing system according to the related art.

[Explanation of symbols]

1: wide-field imaging lens, 2, 17, 602, 80
1, 802, 803, 804 ... television camera,
3, 603, 806: Image processing device, 4: Image input I /
F, 5, 606, 607, 608, 609 image memory, 6 CPU, 7 video output I / F, 8 pan head controller, 9 work memory, 10 program memory, 11
... external I / F, 12 ... data bus, 13, 19, 61
4, 807: video monitor, 16: imaging lens capable of zooming, 18: electric pan head, 51: input image memory, 5
2. Background image memory 53, 54 Work image memory
55: Inter-image calculation result image memory, 56: Binary image memory, 601: Imaging lens, 604: A / D converter, 6
05: memory control unit, 610: image signal output unit, 611: character generation unit, 612: CPU, 613 ... I
/ O control unit, 615 ... calculation unit, 616 ... binarization conversion unit,
805: Image signal switch.

Claims (4)

[Claims] 1. A television camera provided with an imaging lens having a wide field of view for outputting an image signal used for detecting an intruding object, and an image signal mounted on a motor-driven platform for tracking an intruding object and used for image monitoring A television camera equipped with a zoomable imaging lens that outputs an image signal, and an image signal input from the television camera equipped with the wide-field-of-view imaging lens. And an image processing device for outputting a rotation control signal for automatically tracking and capturing an image of an intruding object that has entered a required monitoring location. 2. A television camera provided with an imaging lens having a wide field of view for outputting an image signal used for detecting an intruding object, and a motorized head mounted on a motor-driven platform for tracking an intruding object and monitoring the enlarged image. A television camera equipped with a zoomable imaging lens for outputting an image signal; and an image signal input from the television camera equipped with the imaging lens having a wide field of view. An image processing device that outputs a rotation control signal for moving the zoom lens and a zooming control signal for zooming the zoomable imaging lens by a required amount, and automatically tracks an intruding object that has entered a required monitoring location. An intruding object tracking image processing system characterized by performing automatic zooming and imaging. 3. The image processing device according to claim 1, wherein the image processing device stores an image signal from a television camera having an imaging lens with a wide field of view and an image signal obtained by processing the image signal. An image memory; a CPU for performing processing, calculation, control, management, and the like in the image processing apparatus; a program memory for storing a program for operating the CPU; and position coordinates, detected action history, and the like of the detected intruding object. A work memory for storing data for analyzing the data obtained, and a rotation control signal for rotating the electric head by a required amount based on the calculated data, and a required amount of zooming of the zoomable imaging lens. And a pan head controller for outputting a zooming control signal. 4. An image memory according to claim 1, wherein the image memory stores an input image signal and a background image signal used for detecting an intruding object. A background image memory, two work memories for respectively storing a background image signal and an input image signal for performing an operation for detecting a difference between images, and storing a result of detecting and calculating a difference between the background image and the input image. An intruding object, comprising: an image calculation result image memory for performing an operation, and a binarized image memory for storing a binarized image signal obtained by binarizing a result obtained by detecting and calculating a difference between images. Tracking image processing system.