JP2002352340A - Image monitoring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that a conventional method can not follow a dynamic change such as a sunlight condition of an area to be monitored, resulting in false detection and failure in detection of an intruding object. SOLUTION: An object which moves fast is detected on the basis of a rapidly updated luminance difference between a reference image and a monitored image, an object which moves at a low speed is detected by using an averaging monitored image, and an object which is silhouetted by a silhouette object detecting means is detected when a reflected state is discriminated by an environment analyzing means.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image monitoring apparatus for detecting and monitoring an object to be monitored in a moving background, and particularly to a moving scene such as a wave on a water surface or reflection.
The present invention relates to an image monitoring device that selectively detects and notifies the movement of an intruding object in a moving scene such as the fluctuation of trees.

[0002]

2. Description of the Related Art As an image monitoring apparatus, there have been introduced a large number of apparatuses which, when an intruding object is detected in a monitoring target area, notify an observer of the presence of the intruding object. Conventional image monitoring devices include, for example, an image monitoring device including a camera with a preset function, an infrared sensor, a monitor, and a recording device.

A camera with a preset function is a TV camera capable of controlling the direction of the angle of view and the enlargement ratio, and has a function of storing in advance a combination of the direction of the angle of view and the enlargement ratio at a plurality of locations and selecting the combination at any time. is there. The infrared sensor is installed with the infrared irradiating part and the light receiving part facing each other, and detects the intruding object by detecting that the intruding object passes between the irradiating part and the light receiving part and blocks the infrared ray.

An infrared sensor is provided at one or several places in the monitored area, and the direction of the angle of view and the magnification of the camera with the preset function are adjusted to the position of the infrared sensor. When one of the infrared sensors detects an intruding object, the direction of the angle of view and the magnification are selected in conjunction with the detection, and the image captured by the camera with the preset function is recorded by the recording device.

Further, in the invention described in Japanese Patent Application Laid-Open No. 2000-209571, an intrusion object is detected by an infrared sensor and an image including the intrusion object is detected by using an infrared sensor to detect an intruding object. The angle of view and the enlargement ratio of the camera with the preset function are determined according to the change in luminance by processing in the image processing device.

Further, according to the invention described in Japanese Patent Application Laid-Open No. H11-45379, a camera, a recording device, a monitor,
The image processing apparatus processes an image from a camera, detects a movement in the monitoring target area, records the moving portion as an intruding object, or notifies a monitoring person.

In the invention described in Japanese Patent Application Laid-Open No. H11-355564, an image of a monitoring target area where no intruding object exists is held as a reference image, and the luminance of the current image of the monitoring target area and the luminance of the reference image are compared. The part that produces a difference in luminance is detected as an intruding object. Further, the angle of view and the magnification of the camera are determined so that the position of the intruding object is at the center.

[0008]

However, according to the conventional image monitoring apparatus described above and the invention described in Japanese Patent Application Laid-Open No. 2000-209571, an intruding object is detected only at a place where an infrared sensor is installed in a monitoring target area. The first problem is that detection is possible, and an intruding object cannot be detected in other places.

[0009] Further, in a scene where the background is moving, such as waves and reflections on the water surface and branches and leaves of trees, Japanese Patent Application Laid-Open No. H11-45379.
When the invention described in Japanese Patent Application Laid-Open No. H11-355564 and Japanese Patent Application Laid-Open No. 11-355564 are used, there is a second problem that even when no intruding object exists, these movements in the monitoring target area are erroneously detected as an intruding object.

[0010] By the way, when a ship sails on the sea, a wake occurs behind the ship. When a ship is fast, the wake is noticeable and can be several times the size of the ship. When the inventions described in JP-A-11-45379 and JP-A-11-355768 are applied in such a situation, the wake of the ship is captured as an intruding object, and only the wake is recorded in the recorded image. There is a third challenge.

[0011] The object of the present invention is to create a wave or reflection on the water surface,
It is an object of the present invention to provide an image monitoring device capable of detecting an intruding object with high accuracy without erroneously detecting a motion in a background such as a branch of a tree.

Another object of the present invention is to provide an image processing apparatus capable of accurately measuring the coordinates of the presence of an intruding object and pointing the angle of view of the camera to the coordinates.

[0013]

As a first aspect of the present invention for solving the first problem, an image of a monitoring target area is processed, and an intruding object candidate is determined based on a local luminance difference from a reference image. Then, the intruding object candidate is tracked over several frames, and after confirming that the intruding object has moved a predetermined distance or more, the intruding object candidate is determined as an intruding object and reported to a supervisor or recorded on a recording device.

With the above configuration, if the whole area to be monitored is set within the angle of view of the television camera, an intruding object can be detected in the whole area to be monitored. In most cases, it is acceptable to notify after the intruding object has moved to some extent, and by adopting the first uniformity, catching the intruding object that moves without reacting to the reciprocating motion of the branches and leaves or the flag or the fluctuation of the water surface Becomes possible.

According to a second aspect of the present invention for solving the second problem, a plurality of reference images are provided for each detection target as the reference image, in addition to the same configuration as the first aspect. Provided is an image monitoring device characterized by detecting an intruding object based on a local luminance difference between each reference image and a monitoring target region image.

The first reference image of the plurality of reference images is a reference image which is updated sufficiently quickly using the monitoring target region image from one frame to several frames in the past, and also incorporates fluctuations of waves and trees. Since this is a reference image, it is unlikely that a luminance difference due to fluctuations of waves and trees will occur, and it is unlikely that these waves and fluctuations of trees will be erroneously detected as intruding objects. A luminance difference occurs between the image and the image, so that the intruding object can be detected.

However, when an intruding object enters at an extremely low speed, the intruding object is also captured in the first reference image, like waves and trees that fluctuate, and there is a problem that the intruding object at a low speed cannot be captured. In order to solve this problem, a second reference image is used. The second reference image is obtained by averaging several past frames with a predetermined weight. When several frames are added and averaged, fluctuations of waves and trees are averaged out, and temporal luminance fluctuations are canceled out, while stationary rocks and buildings become sharp reference images with good contrast.

Further, when a surveillance image is averaged with a predetermined weight from the past several frames to the current frame of the image of the surveillance area, the temporal fluctuation of the luminance due to the fluctuation of the waves and trees is canceled out, but the moving speed is low. A moving or stationary intruding object results in a sharp monitoring image with good contrast. When the local luminance difference between the second reference image and the monitoring image based on the averaging is obtained, an intruding object can be detected without catching fluctuations of waves and trees.

However, an object moving at a high speed tends to be blurred in the monitoring image based on the averaging and is hard to detect. Therefore, it is desirable to detect an intruding object higher than a certain speed in the first reference image and detect an intruding object lower than the speed as the second reference image.

A third aspect of the present invention for solving the second problem is characterized in that a reflected light detecting means and a silhouette object detecting means are provided as additions to the same configuration as the second aspect. .

When the luminance of the area detected as the intruder object candidate is higher than a predetermined luminance, the reflected light detection means determines that the intruder object candidate is a reflection of the sun or the like, that is, reflected light, and determines that the intruder object candidate is reflected light. The object detection processing based on the first reference image and the second reference image is temporarily invalidated locally for the region. In other words, the area of the intruder object candidate area is temporarily masked for the object detection processing.

When an intruding object is present in the area where the reflected light exists, as long as the intruding object is not a powerful luminous body, a silhouette state is obtained in which the object portion is always painted black. Therefore, if the intruding object in the silhouette state is detected by the silhouette object detecting means, even if there is an intruding object in the area where the object detection processing based on the first reference image and the second reference image is temporarily invalidated, Can be detected, and the reflected light is not erroneously detected as an intruding object.

According to the third aspect, when an intense light source such as the sun is reflected on the water surface, the reflected light moves in accordance with the fluctuation of the wave, and a local light exists between the first reference image and the monitoring image. It is possible to solve the problem that a difference occurs and an erroneous detection occurs.

As a fourth aspect of the present invention for solving the third problem, according to any one of the first to third aspects, when tracking an intruding object candidate, the speed or the speed of the intruding object candidate is determined. Measures the moving direction, notifies the observer of the part including the contour on the moving direction side of the intruding object candidate as the position of the intruding object candidate, records it, matches the view angle of another camera with the magnification, and It is characterized by doing. With the above tracking process, even if a wake remains after the intruding object, the intruding object itself can be accurately caught.

A moving image obtained from a camera is composed of a series of still images, and in the present specification, each still image which is a series of images is referred to as a frame. A moving image of a standard NTSC television camera is composed of about 30 frames per second.

[0026]

Embodiments of the present invention will be described with reference to the drawings. In the following, a description will be given of an intruder monitoring apparatus at sea as an example, but the present invention is not limited to this.

FIG. 13 shows an example of an embodiment of the image monitoring apparatus according to the present invention. In FIG. 13, a camera 1301 is attached to a support 1303 fixed to a quay or a seaside building 1304.
And a camera 1302. The camera 1301 can be changed in direction and magnification by the angle-of-view adjusting device 1306. The angle of view of the camera 1302 is set so as to cover the entire monitoring target area.

An image of the camera 1301 is displayed on a monitor 1305.
Can be visually observed by the observer, and can be recorded on the recording device 1308. Although not shown, a man-machine interface exists on the angle-of-view adjusting device 1306, the recording device 1308, and the monitor 1305, and the monitor manually operates the camera 13 via the man-machine interface.
By controlling the angle of view 01, the monitoring target area can be visually observed via the monitor 1305. The video of the camera 1301 can be recorded on the recording device 1308 as needed.

When the image processing apparatus of the present invention is set to the automatic mode, the image processing apparatus 1305 processes the image of the camera 1302 and detects, for example, if the intruding ship 1312 exists in the monitoring target area, The angle of view and the magnification of the camera 1301 are automatically controlled via the angle-of-view adjusting device 1306 according to the detected position.

The image processing device 1305 is a recording device 1308
To record the image of the camera 1301,
A monitoring member is notified via the notification device 1307. The monitoring means is notified by the notification means 1307. As a method of notifying, a method such as sound or blinking of light can be used.

A monitor 1310 is provided so that the camera 1
The image of the camera 302 may be recorded, or the recording device 1308 may be provided so that the image of the camera 1302 can be recorded. Also, the recording device 1308 is always recorded using a device having a function capable of overwriting recording,
If the recording operation is stopped at the output of the image processing device 1305,
An image from before the intruding ship 1312 intrudes to immediately after it intrudes can be left.

The cameras 1301 and 1302 do not necessarily need to be installed on the same support.

By adopting the above configuration, if the whole area to be monitored is set to the angle of view of the television camera, there is an effect that an intruding object can be detected in the whole area to be monitored. In most cases, it is OK to notify after the intruding object has moved to some extent, and by adopting the above-mentioned mode, it is possible to catch the moving intruding object without responding to the reciprocating motion of the branches and leaves or the flag, and the fluctuation of the water surface There is an effect that becomes possible.

Next, a description will be given of a second embodiment of the image monitoring apparatus according to the present invention. The second embodiment of the image monitoring apparatus of the present invention has the same configuration as that of the first embodiment, and the configuration of the image processing apparatus 1305 includes an image input unit 101 and a time-series image holding unit 102 as shown in FIG. And the averaging image creating means 103
High-speed and normal-speed moving object detection means 104, silhouette object detection means 105, low-speed and stop object detection means 106, detection object selection means 107, environment analysis means 108, object tracking means 109, It comprises an object determining means 110 and a display means 111.

The image input means 101 takes in the image of the camera 1302 and performs analog / digital conversion. The time-series image holding unit 102 holds a plurality of time-series images. The averaging image creating unit 103 creates an added image by averaging the images of the time-series image holding unit 102 with a predetermined weight.
The high-speed and normal-speed moving object detection unit 104 creates and stores a fluctuation-based reference image from the time-series image of the time-series image holding unit 102.

The fluctuation-added reference image is a reference image which is updated by taking in the fluctuation of the background such as a wave or a grove, and is a reference image from which a change accompanied by a luminance fluctuation larger than the fluctuation of the background is excluded. High-speed and normal-speed moving object detecting means 104
Outputs, as an intruding object candidate, an area in which the luminance difference between the latest image held by the time-series image holding means and the fluctuation-containing background image is equal to or larger than a predetermined value. The output format may be binary image data obtained by binarizing an intruding object candidate area and other areas, or coordinates of a circumscribed rectangle of the intruding object candidate area.

High-speed and normal-speed moving object detecting means 10
No. 4 classifies into a part with a strong fluctuation and a part with a small fluctuation from the distribution of the luminance difference between the addition average image of the addition average image creating means 103 and the latest image of the time-series image holding means 102. Although the intruding object candidate is detected by using a predetermined value as a threshold value for the luminance difference between the latest image and the background image with fluctuation, the threshold value is made variable in a portion where the fluctuation is severe and a portion where the fluctuation is not severe. By increasing the threshold value in a severe part, the possibility of erroneous detection of an intruding object due to fluctuations of waves or trees can be reduced. But this is not required.

By the way, if an intruding object exists between the camera and the light source such as the sun, the intruding object becomes a silhouette. At this time, the image of the camera has a high-luminance area reflected by the light source, and a silhouette of the intruding object is seen as a low-luminance area within the high-luminance area.

The silhouette object detecting means 105 has a function of detecting the silhouetted intruding object. The silhouette object detection means 105 is an addition average image creation means 103
In the addition average image, a region having a luminance threshold or higher is regarded as a reflection region, and a region having a luminance detection threshold or less in the latest reflection image of the time-series image holding unit 102 is defined as a reflection region. It is detected as a silhouetted intruding object candidate.

The low-speed object and stationary object detecting means 106 detects an area of the averaging image of the averaging image generating means 103 where the spatial luminance change is large as a candidate for a low-speed object or a stationary object.

The three object candidate detecting means described above, that is, the moving object detecting means 104 at a high speed and a normal speed, the silhouette object detecting means 105, and the low speed object and the stationary object detecting means 106 are used for monitoring the moving speed of the intruding object. Since the intruding object detection performance changes depending on the presence or absence of reflection of the target area, an effective monitoring device can be obtained by using these in combination.

The detection object selection means 107 integrates the detection results of the high-speed and normal-speed moving object detection means 104, the silhouette object detection means 105, and the low-speed and stationary object detection means 106, respectively.

The image monitoring apparatus of the present invention can set the function of the detection object selection means 107 according to the operating environment. For example, in an operation of arranging a monitor dedicated to monitoring work, the detected object selecting means 107 is manually operated at the discretion of the monitor,
High-speed and normal-speed moving object detection means 104
The detection results by the silhouette object detection means 105 and the low-speed object / stop object detection means 106 can be invalidated or validated. At this time, the effective detected object is in the state of the logical sum of the results of each method.

Furthermore, the environment analyzing means 108 may be provided to determine the environment of the monitoring target area, and the detection object selecting means 107 may automatically switch the validity / invalidity of each detection result.

The display unit 111 displays the operation status of the environment analysis unit 108.
The selection result of the detection object selection means 107 can be displayed. When the observer manually operates the detection object selection unit 107, the environment analysis result displayed by the display unit 111 can be referred to.

The object tracking means 109 is the detection object selecting means 1
The detection object validated in 07 is tracked over a plurality of frames. The intruding object determining means 110 is the object tracking means 109
It is determined whether the tracking object is an intruding object or noise due to fluctuation or the like on the basis of the tracking result, and the result is used as an output of the image processing apparatus 1305.

By adopting the above configuration, reciprocal movement of branches and leaves and flags, and malfunction due to fluctuations in the water surface can be greatly suppressed. In addition, intruding objects having less movement than these background fluctuations and backlighting , The intruding object in the silhouette state can be caught with high sensitivity. Furthermore, environmental analysis means 1
By displaying the analysis result of 08, there is an effect that the operation status can be easily grasped and monitored by the monitor.

The monitoring device of the present invention can be realized by a computer having an image acquisition interface and a signal output interface, and the processing flow is as shown in FIG.

In FIG. 2, each parameter and a memory area for storing an image are initialized (step 1). As a result, malfunction can be prevented, and in addition, the operation for detecting an intruding object is started immediately after the start of the monitoring cycle. After the start of the monitoring cycle, it is checked whether or not there is a termination instruction interrupt. If there is a termination interrupt, the process is terminated. If there is no termination interrupt, the process proceeds to step 3 (step 2).

An image of the monitoring target area is acquired via the image acquisition interface (Step 3). While retaining the required number of time-series images, updating is performed with the images acquired in step 3 (step 4). The averaging image of the time-series images up to the present is updated (step 5).

A high-speed ship valid flag, a silhouette valid flag, and a low-speed ship valid flag are provided, and valid or invalid is set in advance in the initial setting of step 1. If the high-speed ship valid flag is valid, a moving object detection process of the high-speed ship and the normal speed is performed (step 6). If the silhouette valid flag is valid, a silhouette object detection process is performed (step 7). If the low-speed ship valid flag is valid, low-speed and stationary object detection processing is performed (step 8). The processing of steps 6 to 8 may be performed in any order without being limited to the order of FIG.

In the environment analysis processing, it is determined from the luminance of the average added image and the size and the number of the intruding object candidates in steps 6 to 8 whether or not the rough pattern of the waves, the standing of the white waves, and the presence or absence of the reflection are determined in steps 6 to 8. Is set to "valid" or "invalid" of the valid flag, and the level of the threshold value for detecting the intruding object candidate in steps 6 to 8 is controlled (step 9). These flags and threshold values are effective when the image is acquired again in the next monitoring cycle, that is, in step 3, and steps 6 to 8 are executed.

As an example of the environment analysis in step 9, when the average luminance of a partial area of the averaging image is larger than a predetermined value, it is determined that the camera 1302 is in a backlight state facing the light source,
The detection method in step 6 that is likely to be erroneously detected by reflection is temporarily stopped, or the sensitivity is increased by increasing the detection threshold.

In such a backlight state, the sensitivity of the detection method in step 7 is increased because the intruding object has a silhouette shape. If the detection method of step 6 does not reach the determination of the intruding object, that is, if there is a lot of noise, it is determined that the waves are rough and many white waves are standing, and the sensitivity of step 6 is increased. While reducing the possibility of false detection by weakening, the sensitivity of the processing for detecting a low-speed ship in step 8 to detect an object hidden at a low speed and invading at a low speed is maintained.

In the object tracking process (step 10), the positions and characteristics of the intruding object candidates detected in the previous monitoring cycle are held in advance, and the positions and characteristics of the intruding object candidates detected in the current monitoring cycle are collated. The movement status of the intruding object candidate is measured. Such matching of the intruding object candidate during the monitoring cycle is performed, for example, when one or more intruding object candidates are detected in each monitoring cycle, the distance at which the intruding object candidate is located between the monitoring cycles is calculated. Then, the nearest intruder candidate is regarded as the movement of the same intruder candidate.

Further, when the size of the intruding object candidate differs greatly between monitoring cycles and when the distance is large, it can be excluded from the candidates for collation. Further, the similarity of the luminance distribution of the intruding object candidate can be measured and collated by pattern matching processing. In addition, it is also possible to perform collation so that a change in speed up to the previous frame is minimized.

In the display process (step 11), the result of the environment analysis of step 9 is displayed as necessary, the operation status of steps 6 to 8 is displayed, and if an intruding object exists in the monitoring target area, It can be displayed in a frame. In the intruding object determination process (step 12), it is determined as an intruding object from the tracking status of the intruding object candidate,
Angle of view adjustment device 1306 through signal output interface
To cause the camera 1301 to follow, the notifying unit 1307 to notify the observer, and the recording device 1308 to record the image of the intruding object.

Next, in step 2, after checking for an end interrupt, the process reaches the image acquisition in step 3, and the monitoring cycle is repeated.

The above is a computer-based image processing apparatus 130.
5 has the effect that the intruding object can be always detected using the latest calculation speed of the personal computer.

Next, "moving object detecting means 104 at high speed and normal speed" is abbreviated as detecting means 104.
The first embodiment will be described in detail with reference to FIG.

In FIG. 3, a detecting means 104 includes a fluctuation amount measuring means 301 and a fluctuation reference image creating means 302.
The detection invalid area setting means 303 and the object detection means 30
4 and detection object valid / invalid determination means 305. The basic principle is that an image of a monitoring target area where no intruding object exists is retained as a reference image, and a change in luminance of the image of the monitoring target area caused by the intrusion of the object is detected to detect the intruding object.

However, in order to prevent erroneous detection under a background such as a wave or trees, it is necessary to incorporate the fluctuation into the reference image. In addition, since the fluctuation amount to be captured changes every moment, it is necessary to measure this change and automatically adjust the fluctuation capturing amount.

The detecting means 104 has a configuration for realizing this, and the fluctuation amount measuring means 301 measures the luminance fluctuation amount of the fluctuation area in several frames up to the latest frame of the time-series image holding means 102.

As an example of measuring the fluctuation amount, fluctuation amount observation windows 401 to 404 are provided as shown in FIG. FIG.
FIG. 4 is an ocean image captured by the camera 1302, and FIG.
(a) is an image of the monitoring target area of the latest frame, and FIG.
(b) is an image of the monitoring target area several frames before.

For example, fluctuation amount observation windows 401 to 404 are provided at four positions to measure the fluctuation amount of the luminance due to the fluctuation wave.
Is provided. The brightness in the fluctuation observation window 401 fluctuates due to the fluctuation of the wave, and FIG. 4A shows the latest frame and FIG.
FIG. 5A shows an image obtained by calculating a luminance difference in a several frame past image and binarizing the image with a predetermined threshold value.

When the state of the wave fluctuation is large, the luminance change for each frame is large. Therefore, when binarization is performed at a predetermined threshold value, the area of the difference is larger than that of FIG. become that way. When the state of wave fluctuation is small, the luminance change for each frame is small. Therefore, when binarization is performed at the predetermined threshold value, the area of the difference is smaller than that in FIG. 5A, for example, as shown in FIG. 5C. become.

As can be seen from the above description, when the threshold value is fixed, the area of the binarized object increases as the fluctuation of the wave increases, as shown in FIG. 5B, and the area decreases as the fluctuation of the wave decreases. The area of the binarized object becomes smaller as shown in FIG.

The area of the binarized object is compared with the two fixed numerical values. If the area of the binarized object is larger than both of the two fixed numerical values, the binarization threshold is increased, and If the area is smaller than both of the two fixed numerical values, the binarization threshold is reduced to keep the area of the binary object within a certain range, and to increase the threshold according to the magnitude of the wave fluctuation. Can be varied.

That is, when the fluctuation of the wave becomes large, the threshold value becomes large, and as a result, the part where the luminance between the frames greatly changes becomes difficult to be detected as a binary object. As the fluctuation of the wave becomes smaller, the threshold value becomes smaller, and as a result, even a portion where there is little change in luminance between frames is easily detected as a binarized object.

Here, it is assumed that the output of the fluctuation amount measuring means 301 is a binarization threshold value for making the area of the binarized object in the fluctuation amount observation windows 401 to 404 close to a predetermined value. When different threshold values are obtained from the fluctuation amount observation windows 401 to 404, the average value of the four threshold values can be used as the final threshold value, or the median value can be used. Hereinafter, this is referred to as a fluctuation threshold.

The fluctuation reference image creating means 302 creates a fluctuation reference image using the time series image of the time series image holding means 102. Specifically, the initial settings (step
In 1), the latest image acquired by the camera 1302 is set as the fluctuation-containing reference image. For each subsequent monitoring cycle, a brightness difference image between the latest image of the time-series image holding means 102 and the fluctuation-containing reference image is created and binarized by the fluctuation threshold. The binary object detected at this time is the fluctuation amount observation window 4.
This is an area in which a luminance fluctuation larger than the fluctuation of the waves in 01 to 404 occurs.

On the other hand, the fluctuation amount observation window 401 outside the region
To 404, the fluctuation of the luminance is smaller than the fluctuation of the wave in the area 404. In this area only, the reference image with fluctuation is updated with the luminance distribution of the latest image.

By the above processing, the reference image with fluctuations becomes a reference image with fluctuations in the fluctuation amount observation windows 401 to 404 that captures fluctuations smaller than the fluctuations of the waves.

The object detecting means 304 includes the latest image of the time-series image holding means 102 and the reference image forming means 302 with fluctuation.
A luminance difference image with respect to the reference image is generated and binarized at a predetermined threshold value, so that a portion where a luminance fluctuation larger than the fluctuation of the wave in the fluctuation amount observation windows 401 to 404 is generated is converted into a binarized object. Can be detected as

The detection object validity / invalidity determination means 305 sets a circumscribed rectangle of the binarized object, measures the size, and outputs a binarized object in a predetermined range as an intruding object candidate. By adopting the above configuration, it is possible to automatically adjust in response to a change in the amount of fluctuation of the background, and to achieve an image monitoring apparatus with less erroneous detection and high sensitivity.

High speed and normal speed moving object detecting means 10
The fourth embodiment has a configuration in which a detection invalid area setting means 303 is added to the first embodiment. The detection invalid area setting means 303 determines a local area within the monitoring target area based on the latest image of the time-series image holding means 102, the average image of the addition average creation means 103, and the noise occurrence state of the previous monitoring cycle of the detection object valid / invalid determination means. By judging the roughness of the wave and the presence or absence of reflection, the binarization threshold value of the object detection means 304 can be increased in the rough portion of the wave to prevent erroneous detection.

Further, by setting so as not to completely detect a part having a reflection, the occurrence of erroneous detection is prevented.
The detection of the intruding object into the illuminated area can be supplemented by the silhouette object detecting means 105.

FIG. 6 shows an operation example of the detection invalid area setting means 303. FIG. 6A shows a screen of the monitoring target area.
1 is a quay part without fluctuation, and the other part is a wave fluctuation part. When the image of the monitoring target area is processed by the averaging image generating means 103 to generate an averaging image, the area 601 has the same luminance distribution as the corresponding portion of the monitoring target area, and the part other than the area 601 fluctuates. Is the average luminance of the corresponding part, and if there is no reflection, the luminance becomes almost uniform.

When the luminance difference distribution between the averaging image and the image of the monitored area is accumulated over a plurality of cycles for each monitoring cycle and then binarized with a predetermined threshold value, as shown in FIG. The area 602 corresponding to 601 is black, and the other areas are white. Since the white area is a part in which the fluctuation of luminance due to the fluctuation is severe, the detection threshold of the object detecting means 304 is set high, so that erroneous detection can be prevented.

On the other hand, since the black area has a small luminance variation, the possibility of false alarm can be reduced even if the detection threshold of the object detection means 304 is set low to increase the detection sensitivity. Furthermore, the processing of the object detection unit 304 can be stopped in a white area where the luminance value of the averaging image is higher than a predetermined value as a part where reflection occurs.

By adopting the above configuration, even if there is a portion with a lot of fluctuation or reflection in one screen, the optimal adjustment is performed for each area, so that erroneous detection can be performed without lowering the detection sensitivity. It has the effect of suppressing.

Next, details of the first embodiment of the silhouette object detecting means 105 will be described with reference to FIG.

In FIG. 7, the silhouette object detecting means 1
05 comprises a silhouette area detecting means 701 and an object detecting means 702. In the silhouette area detecting means 701, an area having a predetermined brightness or more is selected in the average image by the average image creating means 103.

Further, within the region having the predetermined luminance or higher, the latest image of the time-series image holding means 102 is binarized using another predetermined luminance as a threshold, and the region having low luminance is set as a silhouette object candidate. The object detecting means 702 is a silhouette object detecting means 1 as an intruding object candidate obtained by converting the silhouette object candidate by the silhouette area detecting means 701 into a silhouette.
Output from 05. By adopting the above configuration, there is an effect that a silhouetted intruding object can be caught regardless of its moving speed.

In some surveillance areas, buildings may be present in the surveillance area. In the backlight state, these are also silhouetted. In the first embodiment of the silhouette object detection means 105, the silhouetted intrusion object candidates are turned into silhouettes. I will.
Therefore, the second embodiment of the silhouette object detecting unit 105 has a configuration in which a reference image creating unit 703 and a detection invalid area setting unit 704 are added to the configuration of the first embodiment.

Before the surveillance operation, when an image without an intruding object is input from the camera 102, a silhouette of the building in the surveillance area is detected as in the first embodiment. The reference image is stored in the reference image creation unit 703 as a reference image. The detection invalid area setting means 704 outputs the area of the building in the reference image held by the reference image setting means 703 as an invalid area.

During the monitoring operation, the object detecting means 702 outputs a silhouetted object obtained by removing the invalid area of the detection invalid area setting means 704 from the silhouetted object by the silhouette area detecting means 701.

By adopting the above configuration, there is an effect that an intruding object can be caught without erroneous detection even when there are buildings and the like in the monitoring area and these are silhouetted.

However, as the light source such as the sun moves over time, the shadow of the building also moves. Therefore, the third embodiment of the silhouette object detection unit 105 has a configuration obtained by adding an update time generation unit 705 to the configuration of the second embodiment.

The update time generation means 705 generates a signal at a predetermined time interval, and only when no intruding object is present, generates a signal as a reference signal.
It is told to 03. When an update signal is received, the reference image creating means 703 outputs the silhouette area detecting means 701 at this time.
Is used as a new reference image, and this is hereinafter referred to as a detection invalid area until the next update.

By adopting the above configuration, there is an effect that an intruding object can be caught without erroneous detection even if the manner of forming the silhouette changes with the movement of the light source.

Next, "low speed and stationary object detecting means 106"
Is abbreviated as object detection means 106, and details of the first embodiment of the object detection means 106 will be described with reference to FIG.

When the surveillance video which is constantly fluctuating like a wave is processed by the averaging image generating means 103, the fluctuation portion of the luminance is temporally averaged, and the brightness gradually changes spatially. Poor image. If the intruding object is stationary in the area, the brightness distribution of the intruding object becomes a clear image with a good signal-to-noise ratio due to the effect of the averaging process. Also, when the intruding object moves at a low speed, the image of the intruding object is blurred, but has a clearer luminance distribution than the constantly fluctuating wave.

The object detecting means 106 is a means for detecting an intruding object moving or stopping at a low speed using this property, and is constituted by a spatial differentiation processing means 801 and an object detecting means 802. The spatial differentiation processing means 801 is a process for detecting a region where the luminance changes suddenly spatially, and performs a difference process between adjacent pixels. A well-known differential operator may be used instead of the difference processing. An object is detected by emphasizing a region where the luminance is largely changed spatially by the above-described method, and then binarizing the region with a predetermined threshold value.

The object detecting means 802 is the spatial differential processing means 8
01 is output from the low-speed and stopped object detecting means 106 as a low-speed or stopped intruding object candidate. By adopting the above configuration, there is an effect that an intruding object having less movement than the fluctuation of the background can be detected.

In some monitoring areas, buildings may be present in the monitoring target area, and in the averaging image, the brightness distribution is also clear and detected as an object candidate. Therefore, the second embodiment of the low-speed and stopped object detection unit 106 has a configuration in which a reference image creation unit 803 and a detection invalid area setting unit 804 are added to the configuration of the first embodiment.

When a video image free of an intruding object is input from the camera 102 prior to the monitoring operation, a greatly changing building in the monitoring area is detected as in the first embodiment. The reference image is stored in the reference image creation unit 803 as a reference image. The detection invalid area setting means 804 outputs the area of the building as an invalid area in the reference image held by the reference image creating means 803.

At the time of monitoring operation, the object detection means 802 outputs the object detected by the spatial differentiation processing means 801 excluding the invalid area of the detection invalid area setting means 804 as a low-speed and stopped object. By adopting the above configuration, there is an effect that even if there is a building or a rock stopped in the detection area, an intruding object can be detected without erroneously detecting these.

The brightness pattern of a building changes with the ebb and flow of waves. Therefore, the third embodiment of the object detection unit 106 has a configuration obtained by adding an update time generation unit 805 to the configuration of the second embodiment. The update time generation means 805 generates a signal at a predetermined time interval, and only when no intruding object is present, generates an update signal as the reference image generation means 8.
It is told to 03. When an update signal is received, the reference image creation unit 803 uses the output of the spatial differentiation processing unit 801 at this time as a new reference image, and sets this as a detection invalid area until the next update. By adopting the above configuration, even if a building or a rock changes with time, an intruding object can be captured without erroneously detecting these.

Next, the operation of the environment analyzing means 108 will be described with reference to FIG. The number of detected objects that have not been selected as detected objects per unit time is counted while being detected by the high-speed and normal-speed moving object detection means 104 (step 1). The count result of step 1 is compared with a predetermined number. If the number of occurrences is larger, the process proceeds to step 2;

High speed and normal speed moving object detecting means 10
4 is that there is a lot of noise when there is a lot of fluctuation such as waves,
Since the reflected light is strong, the information is displayed on the display unit 111 (step 2). When the detection object selection means 107 is in the automatic switching mode, the output results of the high-speed and normal-speed moving object detection means 104 are completely invalidated, the fact is displayed on the display means 111, and the detection invalid area setting means is set. For example, the area of the invalidation area set in 303 is increased (step 2).

If the brightness of the averaged image 103 is high, it is determined that reflection has occurred in a backlight state, and that fact is displayed on the display means 111 (step 2). The number of occurrences per unit time of an object that is not selected as a detected object while being detected by the silhouette object detecting means 105 is counted (step 3). The count result of step 3 is compared with a predetermined number. If the number of occurrences is larger, the process proceeds to step 4;

The reason why there is much noise in the silhouette object detecting means 105 is that the reference image of the silhouette reference image creating means 703 is inappropriate due to the movement of the light source. Display means 111
(Step 4). While being detected by the low-speed and stationary object detection means 106, the number of occurrences of an object not detected as a detected object per unit time is counted (step 5). The count result of step 5 is compared with a predetermined number. If the number of occurrences is larger, the process proceeds to step 6, and if the number is smaller or equal, the process ends.

The reason why there is a lot of noise in the low-speed and stationary object detection means 106 is because the shadow of a building or the like changes with the movement of the light source and the reference image of the reference image creation means 803 is inappropriate. At the same time, the information is updated and displayed on the display unit 111 (step 6).

By adopting the above configuration, the process is automatically adjusted to the process suitable for the current environment to be monitored, and the operation status is displayed. It has the effect of being easy to handle.

Next, the details of the object tracking means 109 will be described. The difference between the object tracking method of the object tracking means 109 and the conventional object tracking method will be described with reference to FIG.

FIG. 10A shows a conventional object tracking method, in which an object 1001 detected in the current frame and an object 1002 detected in the previous frame are displayed on the same screen. The center of the object 1001 or the center of the circumscribed rectangle of the object 1001 is denoted by A, and the center of the object 1002 or the center of the circumscribed rectangle of the object 1002 is denoted by B.
In the conventional object tracking method, the vector AB connecting the center A and the center B is defined as the movement of the object between the frames.

FIG. 10B is a diagram showing the object tracking method of the object detection means 109, and shows the object 10 detected in the current frame.
01 and the object 1002 detected in the previous frame are displayed superimposed on the same screen.

The center of the object 1001 or the object 1001
Is described as A, and the center of the object 1002 or the center of the circumscribed rectangle of the object 1002 is described as B. A vector 1003 is a vector representing the movement of the object up to the previous frame. C indicates the intersection of a line extending from the center A of the object 1001 or the center A of the circumscribed rectangle in parallel with the vector 1003 and the outline of the object 1001 or the outline of the circumscribed rectangle. D is the object 1002
1003 from the center of the circle or the center B of the circumscribed rectangle
, And the intersection of the contour of the object 1002 or the contour of the circumscribed rectangle. At this time,
The vector DC connecting D and C is defined as the movement of the object between the frames.

In the example of FIG. 10, an object 1001 between frames is set.
And 1002 do not change shape like a rigid body,
When tracking a flying object that moves while emitting smoke, such as a rocket or a missile, or a ship that is navigating with a splash of waves, it detects it as an object including smoke and splashes.
A and B are not always the center of the moving object itself,
The conventional method has a problem that the movement of the object cannot be measured correctly.

On the other hand, since the object tracking means 109 always captures the leading end of the moving object in the moving direction, there is a high possibility that the moving object itself is tracked, and the moving direction of the object can be correctly measured. The object tracking means 109 obtains a movement vector by a conventional method since the movement vector 1003 to be referred to cannot be obtained immediately after the start of tracking of the object, but this movement vector or the average of the movement vectors over a plurality of frames is used as the reference vector 1003. , Find the tip of the moving direction of the object,
Measure the movement vector.

The above is the tracking method when there is one detected object. The operation when there are two or more detected objects will be described with reference to FIG.

FIG. 11A shows the case where the detected object moves at a low speed. The objects 1001 and 1101 are the objects detected in the current frame, and the object 1002 is the object detected in the previous frame. E is the center of the detected object 1101 or the center of the circumscribed rectangle, and F is the point where the line segment extending in the direction of the movement vector 1003 intersects the contour of the object 1101.

Since the object is slow, the movement vector 1003 obtained by tracking up to the previous frame is small, and the reliability of the movement direction is low. Therefore, collation of objects between frames is performed by collating objects having a short distance. For example, candidates for the detected object 1002 to move in the next frame are the objects 1001 and 1101. Movement vector 10
The object that matches the direction of 03 is 1001, but if the movement vector 1003 is small, the object having the smallest distance, that is, 1101 is associated with the vector BE or the vector DF.
Is a new movement vector.

On the other hand, FIG. 11B shows a case where the detected object moves at high speed. The objects 1001 and 1102 are the objects detected in the current frame, and the object 1002 is the object detected in the previous frame. is there. G is the motion vector 10 through the center of the detected object 1102 or the center of the circumscribed rectangle.
The intersection of the line segment extended in the direction 03 and the contour of the object 1102. At this time, a vector DG or a vector DC can be considered as a movement vector between frames.

In general, an object moving at high speed cannot change its moving direction rapidly. In consideration of this, the vector DC is selected as having a direction and magnitude close to the movement vector 1003 to the previous frame, and therefore, it is determined that the object 1002 in the previous frame has moved to the object 1001 in the current frame.

Object tracking means 109 for performing the above operation
Is shown in FIG.

The object tracking means 109 is the circumscribed rectangle setting means 1
201, a center coordinate and tip coordinate measuring unit 1202, a movement vector evaluation unit 1203, a movement vector selection unit 1204, a center coordinate, a tip coordinate and a movement vector holding unit 1205, and an object coordinate output unit 1206.

The circumscribed rectangle setting means 1201 processes the binary image of the detected object selection means 107 to set a circumscribed rectangle of the detected object. The circumscribed rectangle is a rectangle that the object inscribes.
The center coordinates and tip coordinates measuring means 1202 determine the center coordinates of the circumscribed rectangle, and also measure the tip coordinates based on the center coordinates, the tip coordinates, and the movement vector up to the previous frame held by the movement vector holding means 1205.

The movement vector evaluation means 1203 can use the center coordinates, the tip coordinates, the center coordinates and the tip coordinates of the circumscribed rectangle of the previous frame, and the center coordinates and the tip coordinates of the current frame held by the movement vector holding means 1205. Calculate all the moving vectors. Moving vector selecting means 1204
Are the center coordinates, the tip coordinates and the movement vector holding means 120
5, an appropriate motion vector is selected according to the magnitude of the motion vector up to the previous frame, that is, the speed of the moving object. The center coordinates, the tip coordinates, and the movement vector up to the current frame are determined from the selected movement vector, and these are held in the center coordinates, the tip coordinates, and the movement vector holding unit 1205.

The object coordinate output means 1206 calculates the coordinates of the moving object in the current frame from the movement vector selected by the movement vector selection means 1204,
09 output. By adopting the above configuration, even when tracking a flying object moving while emitting smoke such as a rocket or a missile or a ship navigating with a splash of waves, the head of the traveling direction can be accurately detected. If the camera is moved based on this, there is an effect that the intruding object can be included in the image without catching the subsequent smoke and splashes.

Next, the intruding object determination means 110 calculates the movement locus of the object by accumulating the change in the coordinates of the object for each frame of the object tracking means 109 and moves the object over a predetermined movement distance. Is determined to be an intruding object, and the notifying unit 1307 is operated to notify the observer, the angle of view adjustment device 1306 is controlled to control the camera 1301, and the recording device 1308 is controlled to record an image. . By adopting the above configuration, there is an effect that an intruding object can be detected and an image of the intruding object can be recorded even if the observer does not always look at the monitor.

The TV camera described above is not limited to a camera in the visible wavelength range, but may be a camera with a wide variety of wavelengths such as a near-infrared camera and a far-infrared camera. For example, if a far-infrared camera is used, a monitoring operation can be performed at night without using an illuminating device.

When a near-infrared camera is used, an image similar to that of a visible monochrome camera can be obtained. When a near-infrared LED illuminator is used, the intensity in the visible region is small, so that it is difficult for an intruder to notice. There is an effect that can be related.

[0125]

According to the first embodiment of the present invention, if the whole area to be monitored is set at the angle of view of the television camera, there is an effect that an intruding object can be detected in the whole area to be monitored.
Moreover, if the intruding object is detected and reported after some movement of the intruding object, it is possible to capture the moving intruding object without reacting to the reciprocating movement of the branches and leaves or the flag or the fluctuation of the water surface. is there.

According to the second embodiment of the present invention, by displaying the analysis result of the environment analysis means, there is an effect that the operation status can be easily grasped by the monitor and the system can be easily handled.

When the second embodiment of the present invention is implemented by a computer, there is an effect that an intruding object can be always detected by using the latest calculation speed of the computer.

According to the first embodiment of the high-speed ship detecting means of the image monitoring apparatus of the present invention, an automatic adjustment can be performed quickly in response to a change in the fluctuation amount of the background, and an image having a small number of erroneous detections and a high sensitivity can be obtained. There is an effect that a monitoring device can be realized.

According to the second embodiment of the high-speed ship detecting means of the image monitoring apparatus of the present invention, even if there is a portion where the fluctuation or reflection is severe in one screen, the optimum adjustment is made for each region. Therefore, there is an effect that erroneous detection is suppressed without lowering the detection sensitivity.

According to the first embodiment of the silhouette object detecting means of the image monitoring apparatus of the present invention, there is an effect that a silhouetted intruding object can be caught without depending on its moving speed.

According to the second embodiment of the silhouette object detecting means of the image monitoring apparatus of the present invention, even if there are buildings and the like in the monitoring area and these are turned into silhouettes, an intruding object can be captured without erroneous detection. This has the effect.

According to the third embodiment of the silhouette object detecting means of the image monitoring apparatus of the present invention, even if the manner of forming the silhouette changes with the movement of the light source, the intruding object can be detected without erroneous detection. There is an effect that it can be.

According to the first embodiment of the low-speed and stationary object detecting means of the image monitoring apparatus of the present invention, there is an effect that an intruding object having less movement than fluctuation of the background can be detected.

According to the second embodiment of the low-speed and stationary object detecting means of the image monitoring apparatus of the present invention, even if there are buildings or rocks that are stopped in the detection area, they can be detected without erroneous detection. Has the effect of being able to detect

According to the third embodiment of the low-speed and stationary object detecting means of the image monitoring apparatus of the present invention, even if a building or a rock changes with time, an intruding object can be detected without erroneous detection. Has the effect.

According to the environment analysis means of the image monitoring apparatus of the present invention, the processing is automatically adjusted to the processing suitable for the current environment to be monitored and the operation status is displayed. There is an effect that it is easy to find the defect at an early stage and to deal with it.

According to the object tracking means of the image monitoring apparatus of the present invention, even when tracking a flying object moving while emitting smoke such as a rocket or a missile or a ship navigating with a splash of waves, the traveling direction can be accurately determined. Therefore, if the camera is moved based on the tracking result, the intruding object can be included in the image without catching the subsequent smoke, splash, and the like.

According to the intruding object determining means of the image monitoring apparatus of the present invention, there is an effect that an intruding object can be detected and an image of the intruding object can be recorded, even if the observer does not always look at the monitor.

[Brief description of the drawings]

FIG. 1 is a diagram showing a second embodiment of the image monitoring apparatus of the present invention.

FIG. 2 is a diagram of processing when a second embodiment of the image monitoring apparatus of the present invention is implemented by a computer.

FIG. 3 is a diagram of a first embodiment of high-speed and normal-speed moving object detection means of the present invention.

FIG. 4 is a diagram showing an example of fluctuation amount measurement by the fluctuation amount measuring means of the present invention.

FIG. 5 is a diagram showing an example of a fluctuation amount measurement window of the fluctuation amount measuring means of the present invention.

FIG. 6 is a diagram showing an example of measurement invalidation of the detection invalid area setting means of the present invention.

FIG. 7 is a diagram illustrating an example of a silhouette object detection unit according to the present invention.

FIG. 8 is a diagram illustrating an example of a low-speed and stationary object detection unit according to the present invention.

FIG. 9 is a diagram illustrating an example of processing of an environment analysis unit.

FIG. 10 is a diagram showing a difference between tracking by the object tracking means of the present invention and conventional tracking.

FIG. 11 is a diagram showing an example of a tracking method of the object tracking means of the present invention.

FIG. 12 is a diagram showing an example of the object tracking means of the present invention.

FIG. 13 is a diagram illustrating an example of an embodiment of an image monitoring device according to the present invention.

[Explanation of symbols]

101: image input means, 102: time-series image holding means,
103: Mean averaging image creation means; 104: High speed and normal speed moving object detecting means; 105: Silhouette object detecting means; 106: Low speed and stationary object detecting means; 107: Detected object selecting means; ... object tracking means, 110 ... intruding object determination means, 111 ... display means.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Chieko Onuma 7-1-1, Omikacho, Hitachi City, Ibaraki Prefecture Inside the Hitachi Research Laboratory, Hitachi, Ltd. (72) Keisuke Nakajima 7-1 Omikamachi, Hitachi City, Ibaraki Prefecture No. 1 Inside Hitachi, Ltd.Hitachi Research Laboratory (72) Inventor Toshimichi Okada 32 Miyukicho, Kodaira-shi, Tokyo Inside Hitachi Kokusai Electric Koganei Plant (72) Inventor Sachi Fujii 32 Miyukicho, Kodaira-shi, Tokyo Stock F-term in Hitachi Kokusai Electric Koganei Plant (reference) FA37 FA62 FA66 FA69 GA02 GA19 GA57 HA03 JA16 KA03

Claims (24)

[Claims] 1. An image input unit, a plurality of object detection units for different moving speeds, a detection object selection unit for integrating detection results obtained by the plurality of object detection methods, and an object selected by the detection object selection unit An image monitoring apparatus comprising: an object tracking unit that tracks over a plurality of frames; and an intruding object determination unit that determines that an intruding object has been detected when the object tracking unit can track over a predetermined distance. 2. An image input means, a plurality of object detecting means for each moving speed, a silhouette object detecting means for detecting a silhouette-like object, and the detection results of the plurality of object detecting means and the silhouette object detecting means are integrated. Detecting object selecting means, an object tracking means for tracking an object selected by the detecting object selecting means over a plurality of frames, and determining that an intruding object has been detected when the object tracking means can track a predetermined distance or more. An image monitoring device comprising an intruding object determination unit. 3. An image input means, an object detecting means for detecting a non-silhouette object, a silhouette object detecting means for detecting a silhouette-like object, and a detecting means for integrating detection results by the object detecting means and the silhouette object detecting means. An object selection unit, an object tracking unit that tracks the object selected by the detection object selection unit over a plurality of frames, and an intruding object that determines that an intruding object has been detected when the object tracking unit can track the object over a predetermined distance. An image monitoring device comprising a determination unit. 4. The method according to claim 1, wherein
An image monitoring apparatus, comprising: an environment analysis unit, wherein the detection object selection unit selects according to an analysis result of the environment analysis unit. 5. An image monitoring apparatus according to claim 4, wherein an analysis result of said environment analysis means is displayed. 6. An image monitoring apparatus according to claim 4, wherein a selection state of said detected object selecting means is displayed. 7. The method according to claim 1, wherein
An image surveillance device, wherein a surveillant operates the detection object selection means to select the object. 8. The method according to claim 1, wherein
A fluctuation amount measuring means for measuring a luminance change of the background image of the monitoring area, an intruding object detecting means for providing a luminance change equal to or greater than the luminance change measured by the fluctuation measuring means, and a luminance change equal to or less than the luminance change measured by the fluctuation measuring means. An image surveillance device, comprising an intruding object detecting means for bringing in. 9. A fluctuation amount measuring means for measuring a fluctuation amount of a background image of a monitoring area, and a fluctuation amount which is promptly updated so as to incorporate fluctuation less than the fluctuation amount measured by the fluctuation amount measuring means. A high-speed and normal-speed moving object detecting means, comprising: a fluctuation image generating means for generating a reference image with fluctuation, and an object detecting means for detecting an area having a luminance difference between the fluctuation-containing reference image and the monitoring image as an object detection. An image monitoring device, comprising: 10. The luminance variation of a background image in a monitoring area according to claim 8 or 9, wherein a luminance variation greater than a luminance variation measured by said fluctuation measuring means is measured for an area that fluctuates more than a predetermined luminance fluctuation. An image monitoring apparatus characterized in that an object detecting means for detecting an intruding object causing the above is invalidated. 11. An object detection means according to claim 4, wherein said object detection is performed by detecting an area having a change greater than a predetermined luminance difference with respect to a reference image of a monitoring area. An image monitoring apparatus, wherein the object detection means measures a size of the detected object and abandons tracking with a detected object smaller than a predetermined size as noise. 12. An environment analyzing means according to claim 11, wherein the number of noises generated by said object detecting means is counted to determine whether or not said object detecting means is operating in an environment in which erroneous detection is likely. An image monitoring device, comprising: 13. The object detecting means according to claim 12, wherein the number of noises generated by the object detecting means using a different method and the number of noises generated by the object detecting means using the different method are counted for each of the object detecting means.
An image monitoring apparatus comprising: an environment analysis unit that determines an object detection method suitable for a current environment from the counting result. 14. A surveillance image according to any one of claims 1 to 13, wherein several frames of the surveillance area are averaged in luminance for each pixel, and an area in which the degree of spatial change in luminance is equal to or greater than a predetermined size is detected. An image monitoring device comprising: a low-speed and stationary object detection unit. 15. A reference image according to claim 14, wherein the reference image is formed by averaging a time-series image of the monitoring area for each pixel over a plurality of frames, and the degree of spatial change of luminance in the reference image is a predetermined size. An image monitoring apparatus comprising: a low-speed and stationary object detection unit that determines the above area as an intruding object detection invalid area. 16. An image monitoring apparatus according to claim 15, further comprising a low-speed and stationary object detecting means for generating said reference image at regular intervals. 17. The image processing apparatus according to claim 1, wherein an averaged image obtained by averaging the time-series images in the monitoring area is held, and an area having a predetermined brightness or more in the averaged image is processed. An image monitoring apparatus, comprising: a silhouette object detection unit that detects an area having a brightness equal to or lower than a predetermined brightness and uses the detected area as an intruding object. 18. The image processing apparatus according to claim 17, wherein, when no intruding object is present, a region having a second predetermined brightness or less is detected for the region having the predetermined brightness or more in the averaging image, and the region is determined as the other region. A silhouette object detecting unit that creates a binary image of an area, holds the binary image as a reference image, and sets an area having the second predetermined brightness or less in the binary image as an intruding object detection invalid area. An image monitoring device characterized by the above-mentioned. 19. An image monitoring apparatus according to claim 18, further comprising a silhouette object detecting means for periodically creating said reference image. 20. The method according to claim 1, wherein a detected object of each frame is collated between frames, and a moving direction and a size of the object between the frames are set as a motion vector. An image monitoring apparatus, comprising: an object tracking unit that selects a portion representing the position of the object at the time of the next inter-frame collation based on the magnitude of the movement vector. 21. The method according to claim 20, wherein when the movement vector is smaller than a predetermined size, the center of a circumscribed rectangle of the object at the time of inter-frame matching is set as a representative position of the object, and the movement vector is smaller than a predetermined size. In the case where the object is large, an object tracking means is provided which sets an intersection of a line drawn from the center of the circumscribed rectangle of the object in the direction of the movement vector and the contour of the object or the outline of the circumscribed rectangle as a representative position of the object. Image monitoring device. 22. An image monitoring apparatus according to claim 1, comprising a computer having an image input interface and an image display interface. 23. An image monitoring apparatus according to claim 1, wherein an infrared camera is used as the TV camera. 24. The image monitoring device according to claim 23, further comprising an infrared irradiation device.