JP4298031B2 - Stagnant detection device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an obstacle detection device used for level crossing monitoring, and more particularly to an obstacle detection device capable of detecting that an obstacle is staying.
[0002]
[Prior art]
An obstacle detection device, which is a conventional security system based on image processing, mainly functions on an entry into the monitored area. For example, if there is an object that enters the monitored area, it issues a warning. It was.
[0003]
[Problems to be solved by the invention]
However, the above-described conventional obstacle detection device is effective for monitoring in, for example, a restricted entry area, but is suitable for monitoring, for example, a stagnant entry such as a railroad crossing. There was a problem that not.
[0004]
The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an obstacle detection apparatus that can monitor an obstacle staying in a designated area, not monitoring an entering object.
[0006]
[Means for Solving the Problems]
The present invention for solving the above-described problems of the conventional example includes a camera that captures an image range including a monitoring target area and outputs the image range as image data, and a target image memory that stores the image data. A reference image memory for storing image data obtained by performing gradation conversion on image data of a reference image captured by a process activated according to a change in ambient brightness detected by the sensor, and residence limit time information And a processing memory for storing information for specifying a detected object detected by comparing the target image memory with the reference image memory, and performing the gradation conversion on the image data of the target image memory. The detected object is recognized from the image data stored in the reference image memory and the retention time of the detected object is stored in the retention time limit information. It is characterized by the fact that it is equipped with an arithmetic unit that judges that the detected object is a stagnant and outputs a signal when the time is exceeded, and can only monitor when an obstacle is stagnating in the monitored area In the monitoring, it is possible to easily evaluate the presence or absence of a stagnant substance.
[0007]
Further, the present invention is characterized in that, in the above stagnant detection apparatus, the gradation conversion process for comparing the image data of the target image memory and the image data of the reference image memory is a binarization process. .
[0008]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the claimed invention will be described with reference to the drawings. The obstacle detection device according to the present invention compares a reference image obtained by capturing an image in a state where no obstacle is present in the monitoring target area and a target image obtained by capturing the image being monitored, and sets a set residence time limit. Since an alarm is generated when an obstacle (stagnation) that stays continuously is detected over a period exceeding, a staying obstacle can be monitored.
[0009]
First, the configuration of the obstacle detection apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing the configuration of an obstacle detection apparatus according to the present invention.
[0010]
The obstacle detection device (this device) of the present invention includes a camera 1, a video decoder 2, a target image memory 3, a reference image memory 4, a processing memory 5, an alarm generator 6, and a secondary storage device. 7, an image transmission device 8, an input unit 9, and a high-speed arithmetic device 10.
[0011]
Next, each part of this apparatus is demonstrated concretely.
The camera 1 is an imaging device that captures an image range including a monitoring target area, and outputs a video signal of a general NTSC (National Television System Committee) or PAL (Phase Alternation by Line).
[0012]
The video decoder 2 demodulates the video signal from the camera 1, converts it to a signal in a format such as RGB or YUV, and further performs A / D conversion to output digital image data.
[0013]
The target image memory 3 is an image memory that stores image data output from the video decoder 2 for one screen.
[0014]
The reference image memory 4 is an image memory that stores image data for one screen. The reference image memory 4 stores image data that has been subjected to gradation conversion processing by the high-speed arithmetic device 10 to be described later. .
[0015]
The input unit 9 inputs a monitoring condition in the obstacle detection device and a reference image photographing request for preparing a reference image, and outputs the request to a high-speed arithmetic device 10 to be described later. Or the like, and may be arranged at a location away from other components.
[0016]
The processing memory 5 stores various kinds of information necessary for the monitoring process. Specifically, the monitoring target area information and the stay limit time information set in advance, and the obstacle information stored in the course of the monitoring process, Is remembered.
[0017]
Here, the monitoring target area information is information that represents the monitoring target area to be monitored for obstacles in the imaging range of the camera 1 by the address ranges of the target image memory 3 and the reference image memory 4.
In the present invention, the monitoring target area information is stored in the processing memory 5 by calculating the address range corresponding to the high speed arithmetic unit 10 with respect to the range setting from the input unit 9.
[0018]
The stay limit time information is time information indicating how long the same obstacle stays in the monitoring target area before generating an alarm. The stay limit time set from the input unit 9 is used for processing. It is stored in the memory 5.
[0019]
The obstacle information is status information such as a position and a size of an object recognized as an obstacle in a monitoring process in the high-speed computing device 10 to be described later, and information indicating the staying time.
[0020]
The alarm generator 6 generates an alarm with sound, light, voice or the like in accordance with an alarm instruction signal from a high-speed arithmetic device 10 described later.
[0021]
The secondary storage device 7 is a removable storage device such as an MO disk, a PD disk, or a PCMCIA card, and stores image data of a target image when an alarm is generated.
[0022]
The image transmission device 8 is a general image transmission device that receives a video signal from the camera 1, performs processing necessary for transmission in accordance with a transmission instruction signal from a high-speed arithmetic device 10 to be described later, and transmits it to a telephone line or the like. is there.
[0023]
The high-speed arithmetic device 10 controls the entire obstacle detection device. Specifically, the reference image photographing process for preparing the reference image, the monitoring condition setting process for setting the monitoring condition, and the obstacle that actually stays The monitoring process for monitoring is performed.
[0024]
Here, the reference image photographing process is started in accordance with a reference image photographing request from the input unit 9 and makes it easy to evaluate the presence / absence of a stagnant object in monitoring the image data stored in the target image memory 3. The gradation conversion processing of the image is performed, and the gradation-converted image data is stored in the reference image memory 4.
[0025]
Typical examples of the gradation conversion processing include edge enhancement processing for detecting and enhancing an image boundary, and binarization processing for binarizing image data with a specific threshold value.
Note that the threshold value and the like in binarization are set in consideration of imaging conditions, changes thereof, detection accuracy, and the like.
[0026]
Note that a sensor that senses ambient brightness may be provided in the obstacle detection device, and the reference image capturing process may be activated in response to a change in brightness.
[0027]
The monitoring condition setting process is started in response to an instruction from the input unit 9. When the staying limit time is input from the input unit 9, the monitoring condition setting process is stored in the processing memory 5 as staying limit time information. When the range of the monitoring target area to be monitored is set, the corresponding address range is calculated and stored in the processing memory 5 as monitoring target area information.
[0028]
The monitoring process is started at regular time intervals, and gradation conversion processing is performed on the image data stored in the target image memory 3. The image data after gradation conversion and the image stored in the reference image memory 4 are processed. Based on the data, the presence / absence of an obstacle is evaluated in the target area stored in the processing memory 5.
[0029]
Here, as the simplest method for evaluating the presence or absence of an obstacle, the difference between the image data in the reference image memory 4 and the image data after gradation conversion of the target image is calculated, and the difference is equal to or greater than a specific value. In addition, there is a method of recognizing as one obstacle when the set exceeds a specific size, but the present invention is not limited to this.
[0030]
When a new obstacle is recognized, status information such as position and size regarding each obstacle is stored in the obstacle information in the processing memory 5 and the residence time is set to zero.
At this time, the recognized obstacle is compared with information about the obstacle already stored, and when it is determined that the obstacle is the same, 1 is added to the residence time of the obstacle, and the residence time is However, when the time stored in the stay limit time information of the processing memory 5 is exceeded, this obstacle is determined as a stay and an alarm process is performed.
[0031]
Note that the technology for individually recognizing a plurality of objects that have entered the monitored area has been already put into practical use. "Measurement and Measurement of Moving Objects" pp.174-pp.239 (second volume), Fuji Techno System Co., Ltd., 1994.
[0032]
Specifically, the alarm processing is to output an alarm instruction signal that causes the alarm generator 6 to generate an alarm, generate an alarm with sound, light, voice, etc., and output the alarm signal to the outside to convey the alarm, In this process, the contents of the image memory 3 are stored in the secondary storage device 7 and a transmission instruction signal for transmitting the image to the image transmission device 8 is further output to transmit the image to the outside.
[0033]
Next, the flow of the monitoring process in the high-speed arithmetic device 10 of this apparatus will be described with reference to FIG. FIG. 2 is a flowchart showing the flow of monitoring processing in the high-speed arithmetic device 10 of the present apparatus. Note that FIG. 2 illustrates an example in which the monitoring process is started at intervals of 1 second.
[0034]
The monitoring processing in this apparatus is started at regular time intervals (unit time: for example, 1 second), and gradation conversion processing is performed on the image data stored in the target image memory 3 (100). Obstacle evaluation processing for evaluating whether there is an obstacle by comparing with the contents is performed (102), whether there is an obstacle is determined (104), and when there is no obstacle (No), the monitoring processing is terminated.
[0035]
On the other hand, when there is an obstacle in the process 104 (Yes), the obstacle information in the processing memory 5 is referred to determine whether it is a new obstacle (106), and if it is new (Yes), The obstacle status information is stored in the obstacle information of the processing memory 5 and information storage for setting the staying time to 0 is performed (108), and the processing returns to the processing 102.
[0036]
On the other hand, when the obstacle is not a new obstacle in the process 106 (No), a unit time (for example, 1) is added to the residence time of the corresponding obstacle in the obstacle information in the processing memory 5 to process the residence time. It is determined whether or not the staying limit time stored in the memory 5 has been exceeded (112). If it has not been exceeded (No), the process returns to the process 102. If exceeded, an alarm process is performed (114). To come back.
[0037]
Next, an example in which the monitoring accuracy is improved by monitoring an obstacle from two different directions will be described with reference to FIG. FIG. 3 is a configuration diagram of an image input portion when an obstacle is monitored from two different directions in the present invention.
[0038]
In the present invention, when an obstacle is monitored from two different directions, two sets of a camera 1, a video decoder 2, a target image memory 3, and a reference image memory 4 are provided as shown in FIG. When a stagnant object is detected in the image taken by the camera 1, it is evaluated whether the stagnant object is an obstacle to issue an alarm from the image from the other camera 1. To come out.
[0039]
For example, the camera 1a captures an image range including the monitoring target area from the horizontal direction, and the camera 1b captures an image range including the monitoring target area from the vertical direction.
Then, when an obstacle similar to a cat as shown in FIG. 3 is detected and stays in the monitoring target area of the image photographed by the camera 1b, the thickness of the obstacle from the image photographed by the camera 1a. Can be evaluated.
[0040]
For example, if you are a cat, you can see from the image taken with the camera 1a that the obstacle has a certain thickness, it is recognized as an obstacle, and an alarm is issued, for example, a coffee stain on the floor. In such a case, it can be seen from the image taken by the camera 1a that the obstacle is not thick, the obstacle is not recognized and no alarm is issued, and the detection accuracy of the obstacle to be alarmed is improved.
[0041]
According to the obstacle detection device of the present invention, an obstacle is detected by comparing a reference image without an obstacle with a target image being monitored, and the obstacle stays over a preset stay limit time. If this is the case, it will be judged as a stagnant thing, a sound, light, sound, etc. alarm will be issued, an alarm signal will be output to the outside, and the image at that time will be stored in the removable secondary storage device 7, or image transmission Since it is transmitted to the outside by the device 8, there is an effect that an alarm can be issued when a stagnant object is monitored and detected, not just an entering object.
[0042]
In the present invention, since the monitoring target area can be set within the image range of the camera 1, the monitoring target area can be narrowed down to increase the speed of the obstacle detection process, and the monitoring target area can be freely changed. There is.
[0043]
Further, according to the present invention, since the obstacle is detected by comparing the reference image and the target image with the gradation-converted images, there is an effect that the obstacle detection process can be speeded up.
[0044]
In addition, by taking an image from two directions in the present invention, it is possible to determine the thickness of an obstacle and to improve the detection accuracy of a stagnant object to be alarmed.
[0046]
【The invention's effect】
According to the present invention, the camera captures an image range including the monitoring target region and outputs the image range as image data, the target image memory stores the image data, and the reference image memory has the ambient brightness detected by the sensor. The image data obtained by performing gradation conversion on the reference image captured by the process activated in response to the change is stored, the processing memory stores the retention time limit information, and the target image memory and the reference image memory are stored. Information for identifying a detected object detected by the compared information is stored, and an arithmetic unit calculates a detected object from the image data obtained by performing gradation conversion on the image data of the target image memory and the image data stored in the reference image memory. When the residence time of the detected object exceeds the time stored in the residence limit time information, the detected object is judged as a stay and outputs a signal. Since it is, and can monitor the obstacle remaining in the monitored area, there is an effect that it is possible to easily evaluate the presence or absence of the retentate during monitoring.
[0047]
Further , according to the present invention, since the gradation conversion process for comparing the image data of the target image memory and the image data of the reference image memory is the above-described obstacle detection apparatus which is a binarization process , the monitoring is performed. In this case, there is an effect that it is possible to easily evaluate the presence / absence of the accumulated matter.
[Brief description of the drawings]
FIG. 1 is a block diagram showing the configuration of an obstacle detection apparatus according to the present invention.
FIG. 2 is a flowchart showing a flow of monitoring processing in the high-speed arithmetic device of the present apparatus.
FIG. 3 is a configuration diagram of an image input portion when an obstacle is monitored from two different directions in the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1, 1a, 1b ... Camera, 2 ... Video decoder, 3 ... Object image memory, 4 ... Reference | standard image memory, 5 ... Processing memory, 6 ... Alarm generator, 7 ... Secondary storage device, 8 ... Image transmission device, 9 ... input unit, 10 ... high speed arithmetic unit

Claims (2)

A camera that captures an image range including the monitoring target area and outputs it as image data;
A target image memory for storing the image data;
A reference image memory for storing image data obtained by performing gradation conversion on image data of a reference image captured by a process activated in accordance with a change in ambient brightness detected by a sensor ;
A memory for processing for storing residence limit time information, and storing information for specifying a detected object detected by comparing the target image memory and the reference image memory;
The detected object is recognized from the image data obtained by performing the gradation conversion on the image data of the target image memory and the image data stored in the reference image memory, and the residence time of the detected object is stored in the residence limit time information. A stagnant detection apparatus comprising: an arithmetic unit that determines that the detected substance is a stagnant and outputs a signal when the detected time is exceeded.   The stagnant detection apparatus according to claim 1, wherein the gradation conversion processing for comparing the image data of the target image memory and the image data of the reference image memory is a binarization processing.

Images