JPH04295975A - Monitoring device - Google Patents

Abstract

PURPOSE:To offer a monitoring device capable of detecting the position of a mobile body normally especially against the change of illuminance in the monitoring device which performs the detection of the mobile body using image data. CONSTITUTION:This device consists of a means 3 which calculates the luminance ratio of two pieces of image data different in point of time, a means 4 which calculates the mean value of the luminance ratio, a means 6 which finds deviation from the mean value and counts the number of picture elements with a value higher than a prescribed value, a means 7 which decides the presence/ absence of the mobile body from the number of picture elements, and a means 8 which calculate the center of the mobile body from information with respect to the picture element number and the presence/absence of the mobile body.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a monitoring device that detects the movement of a moving body using an image input device such as a camera.

0002

2. Description of the Related Art FIG. 4 shows the configuration of a conventional monitoring device.

[0003] In FIG. 4, image input means 1 is composed of a camera or the like. The image memory 2 receives image data from the image input means 1 at sample time T, and the stored data is constantly updated.

Alternatively, the input image of the image input means 1 is multiplied by a weighting coefficient α (0≦α≦1), the image memory data is multiplied by a weighting coefficient (1−α), and the results are stored in the image memory. Store as background image data.

The absolute difference means 3 calculates the absolute difference between the output of the image input means 1 and the output of the image memory 2. The binarization means 4 binarizes the output of the absolute difference means 3 with a predetermined value and converts it to "1".
”, “0” is output. The counting means 5 counts the output “1” of the binarizing means 4. The determining means 6 determines from the output of the counting means 5 whether there is a moving object within the field of view. .

Next, the operation of the above conventional example will be explained. The absolute difference value between the output of the image input means 1 and the image memory 2 takes a non-zero value in a temporally changed portion, that is, a moving portion. Therefore, this difference value is binarized using a predetermined threshold, and 2
The number of pixels that are "1" after conversion is compared with a predetermined value, and if this number of pixels exceeds the predetermined value, it can be determined that the moving object is within the field of view. In this way, the conventional monitoring device described above detects a moving object based on the difference in brightness or the like of image data. (For example, Shokodo "Image Processing Handbook" p3
75-376, see JP-A-62-113279. )

[0007]

[Problems to be Solved by the Invention] However, since the conventional monitoring device described above calculates the difference between the input image data and the image memory, when there is a change in illuminance, the difference value may be non-zero even for background parts that are not moving parts. Therefore, there is a problem in that it is difficult to correctly detect a moving object.

The present invention solves these conventional problems, and aims to provide an excellent monitoring device that can detect the center of movement of a moving object even when there is a change in illuminance. be.

[0009]

[Means for Solving the Problems] In order to achieve the above object, the present invention provides a means for determining the luminance ratio of a minimal region of input image data and image memory output data, and a means for determining the luminance ratio of a small region constituted by the minimal regions. means for calculating the average value; a counting means for calculating the deviation of the brightness from the average value of the brightness ratio and counting and outputting the minimum area where the value exceeds a predetermined value for each small area; The apparatus is provided with means for determining the presence of a moving object.

0010

[Operations] The present invention has the following functions due to the above structure.

As shown in FIG. 5, the brightness ratio R(x, y) of input data and image memory in a small area of a divided image is determined by I1(x, y) of the incident light to each small area of input data and image memory. y), I2(x, y), and the surface reflectance of each target object as r1(x, y), r2(
x, y), it is expressed by (Equation 1).

0012

[Math 1]

In (Equation 1), if there is only a change in illuminance, it can be assumed that r1 (x, y) and r2 (x, y) are equal and that the illuminance is uniform in a small area, so
R(x,y) has a constant value within the small area. On the other hand, if there is a moving object within the small area, r1 (x, y) and r2 (
x, y) are not equal, so the result is R(x, y)
The values of are distributed around the mean value. FIGS. 6(a) and 6(b) are
This is a histogram of the brightness ratio R(x,y) when the illuminance changes in a small area and when there is a moving object. By calculating the average value of R (x, y) and the absolute difference value of R (x, y) and counting the number of pixels for which the value exceeds a predetermined value, it is possible to determine whether the brightness change is due to a change in illuminance or due to a moving object. It can be determined from FIGS. 6(a) and 6(b) whether it is a brightness change.

Regarding calculation of the center of the moving body, if the above-mentioned number of pixels in the small area (i, j) where it is determined that the moving body exists is W(i, j), then from (Equation 2), the center of gravity of the moving body can be calculated from (Equation 2). The position (x, y) can be found. However, (Equation 2) is calculated over the entire screen area.

[0015]

[Math 2]

Furthermore, when there are multiple moving objects on the screen, the small areas of the detected moving objects are connected, and (Equation 2) is calculated for W(i, j) belonging to the connected areas. By doing so, the center position (i, j) of each moving body can be found.

[0017]

Embodiment A first embodiment of the present invention will be described below with reference to FIG.

In FIG. 1, image input means 1 is composed of a camera or the like, and inputs an image at every sampling time T. Image memory 2 stores one field or one frame of image data. The brightness ratio calculation means 3 calculates the output ratio of pixels at the same point in the image input means 1 and the image memory 2, that is, the brightness ratio. The average value calculation means 4 calculates the average value of the output of the brightness ratio calculation means 3 in a small area made up of a plurality of pixels. The delay means 5 synchronizes the timing of the output of the average value calculation means 4 and the output of the brightness ratio calculation means 3. The pixel counting means 6 counts and outputs the number of pixels for which the average value of the brightness ratio in the small area and the absolute difference value between the output of the delay means 5 is larger than a predetermined value. The moving object determination means 7 is the pixel counting means 6
Determine whether a moving object exists in the small area from the output. The moving object center calculating means 8 calculates the center of the moving object from the determination result of the moving object determining means 7 for each small region and the output of the pixel counting means 6.

Next, the operation of the above embodiment will be explained. In the above embodiment, image data is input from the image input means 1 at every sampling time T, and is input to the image memory 2 and the brightness ratio calculation means 3. Further, the image memory 2 outputs the stored image data at the same time as inputting the image data. Therefore, the luminance ratio calculating means 3 has a sampling time T
Image data from a separate location will be input. Although the brightness ratio calculation means 3 calculates a simple ratio of two inputs, an equivalent result can be obtained by logarithmic transformation.

FIG. 3 shows an example of the configuration of the luminance ratio calculating means using logarithmic transformation. Input terminals 33 and 34 correspond to the two input terminals of the brightness ratio calculation means 3. Logarithmic conversion 31, which is a logarithmic conversion means, is a means for performing logarithmic conversion of signals inputted from input terminals 33 and 34, and can be realized by a look-up table. The absolute difference calculation means 32 calculates the difference between the two logarithmically transformed signals, and then calculates the absolute value.

Luminance ratio calculation means 3 obtained as described above
The average value of the output is calculated by the average value calculation means 4 for each small region shown in FIG. 5, and is also input to the delay means 5 at the same time. Therefore, the average value of the brightness ratio and the brightness ratio in each small area of the same screen are input to the pixel counting means 6. The pixel counting means 6 compares the absolute difference value between the brightness ratio within a small area and the average value of the brightness ratios of the small area with a predetermined value, and counts the number of pixels exceeding the predetermined value. The number of pixels in each small area is compared with a predetermined value by the moving object determining means 7, and if it is larger than the predetermined value, it is determined that there is a moving object in that small area.

Next, the position and number of pixels of the small area in which it has been determined that a moving object exists are input to the moving object center calculation means 8. Furthermore, these small areas are connected by 4-connection or 8-connection, and the center position is calculated by calculating the center of gravity of each connected area, and the result is output. Furthermore, the center position of the moving body can be calculated by finding the variance value of the brightness ratio instead of the number of pixels.

Next, regarding the second embodiment of the present invention, FIG.
This will be explained with reference to. In the second embodiment, the image memory 2 in the first embodiment described with reference to FIG. 1 is used as a background image memory 2. For example, it can be applied when the image input means 1 is fixed and the background image is known in advance.

Two methods regarding the background image memory 2 will now be explained. One method is to input an image from the image input means 1 into the background image memory 2 when the moving object is not within the field of view. For example, it can be applied when the image input means 1 is fixed and the background image is known in advance.

As another method, in order to deal with slow changes in the background image, the image input every sample time T is I(nT), and the image in the background image memory 2 is M(nT).
T) and weighting coefficient α (0≦α≦1), there is a method of updating the background image memory 2 according to (Equation 3).

[0026]

[Math 3]

[0027] If the value of α is reduced in (Equation 3),
Even if the input image changes for a short time, the background image memory 2 does not change. On the other hand, if the value of α is increased, the background image memory 2 will change even if the input image changes for a short time.

The configuration, operation, etc. other than the background image memory are as follows.
This is the same as the embodiment. Explanation will be omitted.

In this embodiment, the image sampling times are set at equal intervals to T, but even if the sampling intervals are made unequal, the present invention will not be affected in any way.

Further, in this embodiment, the brightness ratio is calculated sequentially for the input image, but the brightness ratio may be calculated by appropriately thinning out the images.

[0031]

As described above, the present invention has the advantage that, in a method of detecting a moving object from temporal changes in an image, a moving part can be detected even when there is a change in illuminance, and the center of the moving object can also be detected. .

[Brief explanation of drawings]

FIG. 1 is a schematic block diagram of a monitoring device in a first embodiment of the present invention.

FIG. 2 is a schematic block diagram of a monitoring device in a second embodiment of the present invention.

FIG. 3 is a schematic block diagram of the brightness ratio calculation means of the monitoring device in the first and second embodiments.

[Fig. 4] Schematic block diagram of a conventional monitoring device

[Figure 5] Explanatory diagram of small areas in the image area

[Fig. 6] (a) is an explanatory diagram of the change in the number of pixels with respect to the brightness ratio. (b) is an explanatory diagram of the change in the number of pixels with respect to the brightness ratio.

[Explanation of symbols]

1 Image input means 2 Image memory 3 Brightness ratio calculation means 4 Average value calculation means 5 Delay means 6 Pixel counting means 7. Moving object determination means 8 Moving body center calculation means

Claims (5)

[Claims] 1. An image input means for inputting and outputting image data, an image memory for storing image data outputted from the image input means, and an image inputting means for inputting and outputting image data, the output of the image input means and the image stored in the image memory being identical. a brightness ratio calculation means for calculating a ratio of brightness values of a minimum area of a point; an average value calculation means for calculating an average value of brightness ratios in a small area constituted by the minimum area from an output of the brightness ratio calculation means; a delay means for delaying the output of the brightness ratio calculation means; and a pixel counting means for counting and outputting minimal regions where the absolute difference value between the average value of the brightness ratio in the small region and the output of the delay means is larger than a predetermined value. and a moving object determining means for determining the presence of a moving object in the small area based on the output of the pixel counting means. 2. An image input means for inputting and outputting image data, a background image memory for storing a background image, and a luminance of a minimum area at the same point of the output of the image input means and the image stored in the background image memory. a brightness ratio calculation means for calculating a ratio of values; an average value calculation means for calculating an average value in a small area constituted by the minimum area from the output of the brightness ratio calculation means; and delaying the output of the brightness ratio calculation means. a pixel counting means for counting and outputting minimal regions in the small region where the absolute difference value between the average value of the luminance ratio and the output of the delaying means is larger than a predetermined value; A monitoring device comprising a moving object determining means for determining the presence of a moving object in the small area. 3. The brightness ratio calculation means includes a logarithmic conversion means for logarithmically converting the output of the image input means and the image stored in the image memory or the background image memory, and calculates an absolute difference value between the logarithmically transformed output. 3. The monitoring device according to claim 1, comprising absolute difference calculation means. 4. The apparatus according to claim 1, further comprising a moving body center calculating means for connecting small areas determined by the moving body determining means as having a moving body and calculating a moving body center of each connected area. monitoring equipment. 5. A moving object center calculation means for calculating a center of gravity position using as a weight the output of the pixel counting means in each small area where the moving object is determined to exist by the moving object determination means. 2. The monitoring device according to 2.