JP2003208599A - Object recognition device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To avoid danger while an opening/closing means or the like is operated by recognizing an object through an image. <P>SOLUTION: An object recognition device comprises the opening/closing means 2 having a drive part, an image input means 1 for imaging the vicinity of the opening/closing means 2, an illuminating means 3 for illuminating the vicinity of the opening/closing means 2, an illumination control means 4 for controlling the illuminating means 3, an illuminance compensating means 5 for compensating an illuminance variation in the image input from the image input means 1, an object recognizing means 6 for recognizing an object present in the vicinity of the opening/closing means 2 in the image input from the illuminance compensating means 5, and a drive part control means 7 for controlling the drive part of the opening/closing means 2 in dependence on the output of the object recognizing means 6. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object recognizing device for recognizing an object by using an image. The present invention also relates to an object recognition device for an automatic opening / closing device that controls opening / closing devices such as shutters and automatic doors.

[0002]

2. Description of the Related Art Conventionally, this type of object recognition device is applied to an automatic opening / closing device, and has adopted a method of extracting an object existing in the vicinity of the opening / closing device with an image and controlling the opening / closing device. This automatic opening and closing device takes the difference between the background image and the current image, recognizes the difference as an object and controls the opening and closing device, and detects the movement of the object by taking the difference between the previous image and the current image and opening and closing. The switchgear is controlled by identifying whether the switchgear exists in a place where the device becomes an obstacle.

A means for detecting the presence or absence of an object by image recognition in this type of object recognition apparatus is disclosed in Japanese Patent Laid-Open No. 5-141.
There is an object recognition means applied to an automatic opening / closing device as disclosed in Japanese Patent Laid-Open No. 148 and Japanese Patent Laid-Open No. 5-256065.

FIG. 29 is a block diagram of the automatic opening / closing device disclosed in Japanese Patent Laid-Open No. 5-141148. The background image input from the image input means 1 is stored in the background image memory 44,
The current image is stored in the current image memory 45, and the image when it is turned on by the illumination means 3 is stored in the illumination image memory 46. The difference binarized image between the current image and the background image is calculated by the current image difference binarization means 47, and the difference binarized image between the current image and the background image when the illumination is turned on is the illumination image difference 2 The value is calculated by the binarizing means 48, and the difference between the current image difference binarizing means 47 and the illumination image difference binarizing means 48 is calculated by the stereoscopic recognition means 49.
The area determination unit 51 calculates the area difference and the area determination unit 51 calculates the area difference. The abnormality determination means 52 performs the abnormality determination based on the output from the shape determination portion 51, and the output of the abnormality determination means 52 drives the drive portion 54 via the drive portion control means 53 to perform the on / off operation of the opening / closing means. 56 is a memory control means.

When an object exists, the current image difference binarizing means 4
At 7, the object part is extracted. Further, an object including a shadow is extracted from the illumination image difference binarizing means when the illumination is turned on. Therefore, if the object is a three-dimensional object, the illumination image difference 2
By obtaining the difference between the binarization unit 48 and the current image difference binarization unit 47 by the area difference unit 50, it can be seen that the shadow portion exists as a difference and that there is a three-dimensional object.

This method is a method of comparing a three-dimensional object with an object in the absence of illumination in order to extract a shadow formed by illumination, but the background is preserved in a complete situation, and the current image also changes in the environment. It is applied when there is not.

[0007] In addition, the document "Journal of the System Control Information Society, No. 3"
Vol. 8, No. 1 (1994) ”, photometric stereo, moire, texture projection, optical projection method, spatial coding, and other methods have been proposed as three-dimensional image measurement methods. There was a recognition of an object as to how to accurately reproduce 3D under a constant environmental condition.

[0008]

[Patent Document 1] JP-A-5-141148

[Patent Document 2] Japanese Unexamined Patent Publication No. 5-256065

[Non-Patent Document 1] Journal of System Control Information Society, Vol. 38,
Issue 1, 1994

[0009]

However, in the conventional method, for example, in the invention of Japanese Patent Laid-Open No. 5-141148, there is a problem that the shadow cannot be extracted if the background changes.

Normally, when installed outdoors, if the brightness around the switchgear changes due to changes in the weather, the change in brightness may be judged to be the object itself or the movement of the object, resulting in a malfunction. . Further, even when installed indoors, a malfunction may occur due to a change in brightness due to a change in illumination from the outside.

A first object of the present invention is to solve the above problems and to determine the presence or absence of an object according to the degree of change in brightness near the object to be recognized.

The second purpose is to improve the accuracy of data distribution by accumulating data for determining the presence or absence of an object.

A third object is to determine the presence / absence of an object by changing the determination standard according to the number of times of data accumulation, and to determine with less malfunction.

A fourth object is to make a judgment with few malfunctions by using a judgment criterion based on a predetermined difference from normal data for judging the presence or absence of an object.

A fifth object is to find the edge feature of the object in the determination of the presence or absence of the object and make a determination with less malfunction.

A sixth object is to find the end feature of the three-dimensional object to determine the presence or absence of the three-dimensional object, and to make a determination with less malfunction.

A seventh object is to improve the accuracy of edge detection by accumulating data for determining the presence / absence of an object.

An eighth object is to judge the edge feature of the object according to the degree of the peak value of the edge with less malfunction.

A ninth object is to make a judgment with few malfunctions by using a judgment criterion based on the difference between the end feature of the object and the data at the normal time.

The tenth object is to prevent the person from noticing by detecting the object in the invisible region.

[0021]

In order to solve the above first object, the present invention provides, as a first solution means, an illumination means, an image input means for receiving the reflected light of the illumination means, and an illumination means. The presence / absence of an object is determined from a lighting control unit that controls lighting, a signal storage unit that stores a differential signal between images when the lighting control unit controls different lighting states, and a data distribution stored by the signal storage unit. It is configured to have a distribution change detecting means for detecting.

Next, the first comparing means for comparing the two data obtained by the distribution change detecting means obtained when the two illuminating means at different positions are individually turned on, and the first comparing means. A first three-dimensional object determining means for detecting a three-dimensional object by detecting that the compared two data are different is first.
This is a second configuration added to the above configuration.

Further, the signal storage means includes an image memory 1, an image memory 2, a difference calculation section for calculating a difference between the image memory 1 and the image memory 2, and a storage memory for storing the signal calculated by the difference calculation section. The third configuration includes an accumulation state determination unit that determines the signal level accumulated in the accumulation memory and an accumulation control unit that controls the number of times accumulation is performed in the accumulation memory.

The distribution change detecting means detects a first threshold value setting section provided according to the number of times of accumulation of the data accumulated by the signal accumulating means, and data having a value equal to or higher than the first threshold value set by the first threshold value setting section. This is a fourth configuration in which a first object determination unit is added.

The distribution change detecting means is provided with a distribution comparing means for comparing the data distribution accumulated by the signal accumulating means at a normal time with the data distribution accumulated by the signal accumulating means at the time of detection, and a second distribution comparing means. This is a fifth configuration in which a second object determination unit that detects data having a value equal to or greater than a threshold value is added to the first configuration.

Further, between the illumination means, the image input means for receiving the reflected light of the illumination means, the illumination control means for controlling the illumination of the illumination means, and the images when the illumination control means controls different illumination states. Signal storage means for storing the differential signal of
The sixth configuration has a differential operation means for calculating a differential value of the data distribution accumulated by the signal storage means and an edge change detection means for detecting the presence or absence of an object from the data calculated by the differential operation means.

The second comparing means for comparing the two data obtained by the distribution differentiation calculating means obtained when the two illuminating means at different positions are individually turned on and the second comparing means for comparing 2 data A seventh configuration is provided by adding a second three-dimensional object determination means for detecting a three-dimensional object by detecting that the two data are different from each other.

The signal storage means is an image memory 1, an image memory 2, a difference calculation section for calculating a difference between the image memory 1 and the image memory 2, a storage memory for storing the signals calculated by the difference calculation section, and a storage memory. In the eighth configuration, a storage state determination unit for determining the signal level stored in the storage unit and a storage control unit for controlling the number of storages in the storage memory are added to the sixth configuration.

The edge change detection means detects a peak of the data calculated by the differential calculation means, and a third object determination for detecting data whose peak value detected by the peak detection section is a third threshold value or more. And a part are added to the sixth structure as a ninth structure.

The edge change detecting means has a peak comparing section for comparing the data calculated by the differential calculating means at the normal time with the data calculated by the differential calculating means at the time of detection, and a fourth threshold value or more provided in the peak comparing section. Fourth to detect value data
This is a tenth configuration in which an object determination unit is added to the sixth configuration.

An eleventh configuration is obtained by adding the configuration using the illumination means and the image input means in the invisible region to the first to tenth configurations.

[0032]

According to the first aspect of the present invention, the illumination control means controls the illumination means in different states, and the difference signal between the images captured by the image input means in these different states is stored in the signal storage means. The presence or absence of an object is detected by the distribution change detecting means from the data distribution accumulated by the signal accumulating means.

With the second configuration, two different positions are provided.
The two data obtained by the distribution change detecting means obtained when the two illuminating means are individually turned on are compared by the first comparing means, and the two data compared by the first comparing means are different. 1 The three-dimensional object determination means determines.

With the third configuration, in the signal storage means, the difference calculation between the image memory 1 and the image memory 2 is performed by the difference calculation section, and the signal calculated by the difference calculation section is stored in the storage memory.
The signal level accumulated in the accumulating memory is judged by the accumulating state judging section, the number of times of accumulating in the accumulating memory is counted by the accumulating control section, or the accumulating control section is controlled so as to accumulate up to the preset number of times. Data.

According to the fourth configuration, the distribution change detecting means detects, by the first object judging section, data having a value equal to or higher than the threshold value of the first threshold value setting section provided according to the number of times of accumulation of the data accumulated by the signal accumulating means. To determine the presence or absence of an object.

With the fifth configuration, in the distribution change detecting means, the distribution comparing means compares the data distribution accumulated in the signal accumulating means at the normal time with the data distribution accumulated in the signal accumulating means at the time of detection, and in the distribution comparing means. The presence / absence of an object is determined by detecting, by the second object determination unit, data having a value equal to or greater than the provided second threshold value.

According to the sixth configuration, the illumination control means controls the illumination means in different states, and the difference signal between the images captured by the image input means in these different states is accumulated in the signal accumulating means, and the signal accumulating means. The differential value of the data distribution accumulated in 1 is calculated by the differential calculation means, and the presence or absence of an object is detected from the data calculated by the differential calculation means by the edge change detection means to determine the presence or absence of the object.

According to the seventh structure, the two different positions are provided.
The two data obtained by the distribution differentiation operation means obtained when the two illumination means are individually turned on are compared by the second comparison means, and the two data compared by the second comparison means are different. 2 Determined by the three-dimensional object determination means.

According to the fifteenth configuration, in the signal storage means, the difference calculation between the image memory 1 and the image memory 2 is performed by the difference calculation section, and the signal calculated by the difference calculation section is stored in the storage memory.
The signal level accumulated in the accumulation memory is determined by the accumulation state determination unit, and the accumulation control unit counts the number of accumulations in the accumulation memory, or controls the accumulation control unit to accumulate the signal up to a preset number of times. Data.

According to the eighth configuration, in the signal storage means, the difference calculation between the image memory 1 and the image memory 2 is performed by the difference calculation section, and the signal calculated by the difference calculation section is stored in the storage memory.
The signal level accumulated in the accumulation memory is determined by the accumulation state determination unit, and the accumulation control unit counts the number of accumulations in the accumulation memory, or controls the accumulation control unit to accumulate the signal up to a preset number of times. Data.

According to the ninth configuration, the edge change detection means detects the peak of the data calculated by the differential calculation means by the detection peak detection portion, and whether the peak value detected by the peak detection portion is a value equal to or larger than the third threshold value. Is determined by the third object determination unit.

According to the tenth configuration, in the edge change detecting means, the data calculated by the differential calculating means at the normal time and the data calculated by the differential calculating means at the time of detection are compared by the peak comparing section, and provided by the peak comparing section. The fourth object determination unit determines whether the value is a certain threshold value or more.

According to the eleventh configuration, by using the illumination means and the image input means in the invisible area, it is possible to recognize an object invisible.

[0044]

Embodiments of the present invention will be described below with reference to the drawings.

In the first embodiment, an example in which the first structure is applied to a shutter as an automatic opening / closing device will be described.

As shown in FIG. 1, the image input means 1 photographs the vicinity of the opening / closing means 2. In this embodiment, a CCD camera installed from the upper side to the lower side of the automatic opening / closing device is used as the image input means 1.

The illumination means 3 illuminates the vicinity of the opening / closing means 2. In the present embodiment, the automatic opening / closing device is installed from the upper side to the lower side, and the lighting control means 4 controls the turning on / off.

The illuminance correction means 5 inputs the image photographed by the image input means 1 and outputs an image in which the illuminance near the opening / closing means 2 is corrected.

The object recognition means 6 recognizes an object in the vicinity of the opening / closing means 2 from the image output by the illuminance correction means 5 and outputs it to the drive section control means 7.

The drive section control means 7 controls the drive section of the opening / closing means 2 based on the output of the object recognition means 6.

The illuminance correction means 5 has a third configuration, that is, an off-state image storage means 8a for storing an image when the illumination is off, an on-image image storage means 8b for storing an image when the illumination is on, an illuminance measuring means 9, an image difference. Means 10, storage frequency determination means 11,
It is composed of the difference image storage means 12.

The object recognition means 6 is an initial image storage means 13,
The current image storage means 14 and the image comparison means 15 are included.

Next, the operation of the above configuration will be described.

First, the operation of initial setting at the time of installation of the automatic switchgear will be described.

When the power is turned on, the image input means 1 captures an image near the opening / closing means 2. The captured image is sent to the illuminance correction means 5.

The illuminance correction means 5 corrects the input image to an illuminance optimum for the object recognition means 6 to recognize, and outputs it to the object recognition means 6.

The input image is stored in the initial image storage means 13 as the initial image, and the initial setting is completed.

Next, the operation during use will be described.

When the power is turned on, the image input means 1 captures an image near the opening / closing means 2. The captured image is sent to the illuminance correction means 5.

The illuminance correction means 5 corrects the input image to an illuminance optimum for the object recognition means 6 to recognize, and outputs it to the object recognition means 6.

The input image is stored in the current image storage means 14 as the current image.

The image comparison means 15 includes the initial image stored in the initial image storage means 13 at the time of initial setting and the current image storage means 1.
The current image stored in 4 is compared, and if it is determined that there is a difference between the two, it is determined that an object exists near the opening / closing means 2, and if there is no difference between the two, it is determined that no object exists.
The result is output to the drive unit control means 7.

Based on the signal from the object recognizing means 6, the driving section control means 7 transmits a stop signal to the driving section of the opening / closing means 2 when an object exists near the opening / closing means 2, and when the object does not exist. An operation signal is transmitted to the drive unit of the opening / closing means 2.

The illuminance in the vicinity of the opening / closing means 2 changes depending on the installation location, changes in climatic conditions, ambient lighting conditions, and the like. In order to perform image recognition with high accuracy, it is necessary to keep the illumination conditions constant, but the image input means 1 inputs images with various illuminances according to changes in the surrounding environment.
For example, when it is installed outdoors, there is a large difference in illuminance between day and night, and when it is directly illuminated by sunlight due to changes in the weather and when it is cloudy. In the present invention, the illuminance correction means 5 is provided to absorb the difference in illuminance and determine the presence of an object under a certain range of illumination conditions.

When the surroundings are dark, an image with a certain range of illuminance is obtained by being illuminated by the illumination means 3. For example, when illuminated by sunlight, both the light by the illumination means 3 and the sunlight are obtained. Since an image illuminated by is input, an image with a constant illuminance cannot be obtained. However, it is possible to compare the images when the illumination is turned on and when the illumination is turned off, extract the information based on only the illumination from the difference, and create an image corresponding to the one captured under the illumination condition of a certain range.

FIGS. 3, 4, 5, and 6 show image data when a box is placed near the opening / closing means 2 and an image is taken. The illumination means 3 of the switchgear is illuminated from the right side of the figure, and the external illumination is illuminated from the left side of the figure.

When there is no external illumination, the luminance data of the portion indicated by the straight line 1 in FIG. 3 which is an image taken with the illumination of the switchgear turned off is shown in FIG. 7 and the illumination of the switchgear is turned on. FIG. 8 shows luminance data of a portion indicated by a straight line 2 in FIG. 4 which is an image photographed by using the image, and FIG. 9 shows difference data between the data of FIG. 8 and the data of FIG.

Further, when there is an external illumination, the luminance data of the portion indicated by the straight line 3 in FIG. 5, which is an image photographed with the illumination of the switchgear turned off, is shown in FIG. FIG. 11 shows the luminance data of the portion indicated by the straight line 4 in FIG. 6 which is an image photographed with lighting, and FIG.
The difference data of the data of 0 is shown in FIG.

When there is no external illumination, the difference data is almost the same as the data of the image photographed with the illumination turned on, and it can be seen that there is no external influence. On the other hand, when there is another light outside, by subtracting the data of FIG. 10 from the data of the data taken with the light turned off, that is, the data of FIG.
The data shown in FIG. 12 excluding the influence of external illumination is obtained. By repeating the same process and repeating the addition of the data obtained by the difference, almost the same data as in FIG. 9 can be obtained, and an image without the influence of external illumination can be obtained.

The operation of the illuminance correction means 5 will be described.

In response to a signal from the illuminance correction means 5, the illumination control means 4 turns off the illumination means 3 and the image input means 1 photographs the vicinity of the opening / closing means 2. The image input from the image input means 1 is stored in the image storage means 8a when the light is off.

Next, by the signal from the illuminance correction means 5,
The illumination control means 4 turns on the illumination means 3, and the image input means 1 photographs the vicinity of the opening / closing means 2. The image input from the image input means 1 is stored in the image storage means 8b during lighting.

Illuminance measuring means 9 is image storage means 8 when light is off.
a, the illuminance in the vicinity of the opening / closing means 2 is measured from the brightness information of the image stored in the image storage means 8b during lighting.

The image difference means 10 is the image storage means 8 when the light is off.
a, the difference between the images stored in the lighting image storage means 8b is obtained and output to the difference image storage means 12.

Based on the two illuminance information measured by the illuminance measuring means 9, the accumulation number determining means 11 reduces the accumulation number of the difference image data when the difference between the two illuminance information is large, and the difference between the two illuminance information is small. At times, the number of times of storage of the difference image data is increased, and the determined number of times of storage is output to the difference image storage means 12.

The difference image accumulation means 12 is based on the accumulation times determined by the accumulation times determination means 11
The difference image from is additionally stored.

As described above, according to this embodiment, since image data is created by removing the difference in the surrounding illumination conditions and the image recognition is performed, a safe shutter that does not malfunction due to changes in the illumination conditions can be realized.

Next, in the second embodiment, an example in which the second structure is applied to a shutter as an automatic opening / closing device will be described.

As shown in FIG. 2, it is the same as the first embodiment except that the specific background means 16 is provided. In this embodiment, it is a straight line with a constant width drawn on the floor below the shutter.

The first point is that the object recognition means 6 detects the specific background means 16 in the image output by the illuminance correction means 5, recognizes the object on the specific background means 16 and outputs it to the drive unit control means 7. Is different from the embodiment described above.

The object recognition means 6 is an image information extraction means 17,
It comprises an initial image storage means 13, a current image storage means 14, and an image comparison means 15.

When an image is input from the illuminance correction means 5 to the object recognition means 6 at the time of initial setting, the image information extraction means 17
Is extracted from the input image as a feature given by the specific background means 16 and is stored in the initial image storage means 13 as the initial image information. To do.

In use, when an image is input from the illuminance correction means 5 to the object recognition means 6, the image information extraction means 17 draws on the floor as a feature given by the specific background means 16 from the input image. The length, width, and area of the extracted straight line image are extracted and stored in the current image storage means 14 as current image information.

The image comparison means 15 includes the initial image stored in the initial image storage means 13 at the time of initial setting and the current image storage means 1.
4 is compared with the current image stored in 4, and if it is determined that there is a difference between the two, it is determined that an object exists on the specific background means 16, and if there is no difference between the two, there is no object on the specific background means 16. And outputs the result to the drive unit control means 7.

As described above, according to the present embodiment, the feature provided by the specific background means in the image stored at the time of initial setting is compared with the feature provided by the specific background means in the current image. Since the object is recognized automatically, the object can be stably recognized regardless of the shape and type of the object, and the accuracy of recognition of the object near the shutter can be improved.

The third embodiment of the third structure has been described with the structure of the illuminance correction means 5 shown in the first structure.

Next, a fourth embodiment of the present invention will be described with reference to the drawings. In the fourth embodiment, FIG.
As shown in FIG. 3, 1 is an image input unit that receives the reflected light of the illumination unit 3, 4 is an illumination control unit that controls the illumination of the illumination unit 3, and 17 is an illumination control unit 4 that controls different illumination states. Signal accumulating means for accumulating a differential signal between time images,
Reference numeral 18 denotes a distribution change detecting means for detecting the presence or absence of an object from the data distribution accumulated by the signal accumulating means 17.

The signal storage means 17 has an image memory 1 (19), an image memory 2 (20), and
A difference calculation unit 21 that performs a difference calculation between the image memory 1 (19) and the image memory 2 (20), a storage memory 22 that stores the signal calculated by the difference calculation unit 21, and a storage that determines the storage state of the storage memory 22. The state determination unit 23 and a storage control unit 24 that counts the number of times of storage or controls the number of times of storage.

As the ninth configuration of the distribution change detecting means 18, the first threshold setting section 25 and the first threshold setting section 25, which are provided according to the number of times of accumulation of the data accumulated by the signal accumulating means 17, are set. The first object determination unit 26 detects data having a value equal to or greater than the first threshold.

Next, the operation of the structure of the fourth embodiment will be described.

FIG. 14 is a flow chart when the object recognition operation is started. Lighting means 3 by lighting control means 4
The image is turned off at step 101, and the image input means 1 stores at step 102 an image in the state where the illumination is turned off (hereinafter referred to as off image) in the image memory 1 (19) of the signal accumulating means 17. Next, the state in which the illumination means 3 is turned on by the illumination control means 4 is created in step 103, and the image in which the illumination is turned on to the image memory 2 (20) of the signal storage means 17 in step 104 by the image input means 1 ( (Hereinafter referred to as “on image”) is stored.

In step 105, the image memory 1 (19)
The difference calculation between the image memory 2 (20) and the image memory 2 (20) is performed by the difference calculation unit 21, and the calculation result is stored in the storage memory 22 in step 106. In step 107, the accumulation number control unit 11 counts the number of accumulations and checks whether accumulation is performed for a predetermined amount or more. If the accumulated amount is equal to or more than the predetermined amount, the process proceeds to step 109. If the number of accumulations is less than the predetermined number, the process proceeds to step 108. In step 108, the accumulation state determination unit 23 determines the peak value of the image data stored in the storage memory 22. If the peak value is less than or equal to the predetermined value, the process proceeds to step 101 and the image is input again by the image input means with the illumination turned off.

Here, the actual data will be described with reference to FIG. The graph showing the accumulated data is shown in FIG. 15A when the object is not placed. In this graph, the horizontal axis represents position, and the vertical axis represents luminance difference (256 gradations). Illumination means 3a and 3b (marked with asterisks) are arranged on the left and right, and image input means 1 (⊚) is arranged in the center, so that they are overlapped on the graph. Since the illumination means 3a is placed at the left side position (marked with *), the left side shows a slightly higher value. The line a, line b, line c, line d, line e, and line f are stored in the image memory 1 (19) and the image memory (7).
The image brightness difference is added up, and the state where the data accumulation is completed when the number of accumulation times is 6 is shown.

For example, the maximum value is 200 when the number of accumulations is set.
There are various methods such as a method of completing when the average value exceeds / 255, and a method of completing when the average value exceeds 180/255. Therefore, the number of times of accumulation may be determined in advance according to the brightness, or may be determined according to the above-described upper limit value of data. Similarly, FIG. 15B shows the case where there is no object.
And, the case where the illumination means 3b is placed on the right side is shown.

When all the data are accumulated, step 1
At 09, the first object determination unit 14 of the distribution change detection means 18 determines the presence or absence of an object. That is, since the accumulated data is an optical signal having a difference between when there is illumination and when there is no illumination, the light reflected from the floor surface is extracted as an optical signal of a certain level, for example, after several accumulations. For example, when the object Ob is placed on the floor and the illuminating means 3a is adapted to cast a shadow of the object Ob, the shadow portion has no reflection due to the illumination, so that the image input means 1 receives light. This is not the case (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, when there is no object Ob, there is no such shadow, so the accumulated data is the first data in all locations.
When the threshold value is not less than T, the process proceeds to step 110 and the object recognition is completed assuming that there is no object.

Now, in the case where there is an object Ob, FIG.
It will be described together with (c). In the graph of FIG. 15C, the horizontal axis represents the position and the vertical axis represents the luminance difference (256 gradations). Illumination means 3a and 3b (marked with asterisks) are arranged on the left and right, and image input means 1 (⊚) is arranged in the center, so that they are overlapped on the graph. Since the illumination means 3a is placed at the left side position (marked with *), the left side shows a slightly higher value. Line f shows the data when the accumulation is completed. FIG. 15D shows the case where the object Ob is also present.

Here, in FIG. 15C, the object Ob is placed slightly to the right of the center. The illumination means 3a creates a shadow of the object Ob. Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG.
When there is b, the position of the accumulated data is low as a signal due to the shadow. Therefore, since it does not exceed the first threshold value T in all places, the process proceeds to step 111, and the object recognition is completed because there is an object.

FIG. 15D shows the case where there is an object. And, the case where the illumination means 3b is placed on the right side is shown.

The configurations described above are the fourth configuration, the case where the signal accumulating means is the eighth configuration, and the case where the distribution change detecting means is the ninth configuration.
7. The distribution change detecting means 18 can determine the presence or absence of an object, which is the first purpose, if the same function is achieved by using other methods.

In the fourth structure, the lighting control means 4 turns the lighting means 3 on and off, but since the purpose is to create an illuminance difference, the life of the lighting means 3 is extended and other purposes. Therefore, it is okay to stop the off state and illuminate in a weak state, and create a difference in illuminance difference depending on the intensity.

A fifth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 16, in the fifth embodiment, two lighting means 3a and 3b are provided as the lighting means 3 in the fourth structure and are installed at different positions. Further, the first comparing means 2 for comparing the signals of the two distribution change detecting means 18
7 is added to the first three-dimensional object determining means 28 for detecting whether the two data compared by the first comparing means 27 are different and determining whether or not the three-dimensional object is a three-dimensional object.

Next, the operation of the structure of the fifth embodiment will be described.

In FIG. 15C, the object Ob is placed slightly to the right of the center. The illumination means 3a is the object O
It is designed to create a shadow of b. Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG.
When there is b, the position of the accumulated data is low as a signal due to the shadow. Therefore, it does not exceed the first threshold value T in all places. From this, the location below the first threshold T is known, and the value estimated to be the height from the width is measured. At this point, the width is estimated as the value corresponding to the height, but it is simply specified whether a low-reflectance paper-like object is placed or the height of an object with a small brightness difference created by a shadow. I can't.

Next, in FIG. 15 (d), similarly, the illuminating means 3a forms a shadow of the object Ob.
Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG.
When there is b, the position of the accumulated data is low as a signal due to the shadow. Therefore, it does not exceed the first threshold value T in all places. From this, the location below the first threshold T is known, and the value estimated to be the height from the width is measured.

The first comparing means 27 compares the left and right brightness difference portions, that is, the position data less than the threshold value T, and the first three-dimensional object determining means 28 determines whether or not they exist at different positions. If the objects Ob are at different positions, the object Ob is determined as a three-dimensional object.

A sixth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 17, in the sixth embodiment, the opening / closing means 2 and the drive unit 29 for driving the opening / closing means 2 are provided in the fourth configuration.
And a drive unit control means 17 for controlling the drive unit 29 are added.

Next, the operation of the structure of the sixth embodiment will be described.

As described in the operation in the fourth configuration, if the distribution change detecting means 18 determines that the brightness difference distribution data is less than the first threshold value T, there is an object. The image pickup area of the image input means 1 is set to a portion where the end portion of the opening / closing means 2 corresponds to a closed state, and when an object exists at a position that hinders the opening / closing operation of the opening / closing means 2, the object determined previously is not crushed. The drive unit 29 is stopped via the drive control means 30. on the other hand,
The brightness change distribution data is detected by the distribution change detection means 18 as the first threshold value T.
If it is above, there is no object. At that time, the drive unit 29 is driven through the drive control unit 30 to operate the opening / closing unit.

By doing so, safe operation can be performed even when an object is placed under the opening / closing means such as a shutter.

A seventh embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 18, the seventh embodiment has an opening / closing means 2 and a drive unit 29 for driving the opening / closing means 2 in the fifth configuration.
And drive control means 30 for controlling the drive unit 29 are added.

Next, the operation of the structure of the seventh embodiment will be described.

As described in the operation in the fifth configuration, the first
The presence or absence of a three-dimensional object is detected by the three-dimensional object determination means 28. Similar to the fifth embodiment, the object Ob is placed slightly to the right of the center in FIG. 15 (c). The illumination means 3a creates a shadow of the object Ob. Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG.
When there is b, the position of the accumulated data is low as a signal due to the shadow. Therefore, it does not exceed the first threshold value T in all places. From this, the location below the first threshold T is known, and the value estimated to be the height from the width is measured. At this point, the width is estimated as the value corresponding to the height, but it is simply specified whether a low-reflectance paper-like object is placed or the height of an object with a small brightness difference created by a shadow. I can't.

Next, in FIG. 15 (d), similarly, the illuminating means 3a is adapted to cast a shadow of the object Ob.
Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG. 15d, when there is an object Ob, the position of the accumulated data is low as a signal due to the shadow. Therefore, it does not exceed the first threshold value T in all places. First from this
The location below the threshold T is known and the value estimated to be the height from that width is measured.

The first comparing means 27 compares the left and right brightness difference portions, that is, the position data less than the threshold value T, and the first three-dimensional object determining means 28 determines whether or not they exist at different positions. If the objects Ob are at different positions, the object Ob is determined as a three-dimensional object.

Now, when the image pickup area of the image input means 1 is set to a portion where the end portion of the opening / closing means 2 corresponds to the closed state, and a three-dimensional object exists at a position that hinders the opening / closing operation of the opening / closing means 2, it is judged earlier. In order not to crush the three-dimensional object, the drive unit 29 is stopped via the drive control means 30. On the other hand, when the three-dimensional object determination means 15 does not exist, the drive unit 29 is driven via the drive control means 30 to operate the opening / closing means.

By doing so, safe operation can be performed even when a three-dimensional object is placed under the opening / closing means such as a shutter.

The operation of the eighth and ninth embodiments of the present invention has already been described in the portion described in the fourth configuration.

The tenth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 19, in the tenth embodiment, the structure of the distribution change detecting means 18 is set to the normal signal storing means 17
The distribution comparison unit 31 that compares the data distribution accumulated in step S1 with the data distribution accumulated by the signal storage unit 17 at the time of detection, and the threshold value T2 for comparing the data in the distribution comparison unit 31 are set in the second threshold value setting unit 32. , And a second object determination unit 33 that detects whether or not the data has a value of a threshold value T2 or more.

Next, the operation of the structure of the tenth embodiment will be described.

FIG. 20 is a diagram showing comparison data at the time of normal and at the time of detection. The vertical axis represents the brightness difference and the horizontal axis represents the position. Only (c) of FIG. 20 has a gray scale of -127 to +127 because the vertical axis represents the comparison of the brightness differences.

As shown in FIG. 20 (a), when it is considered to be normal in advance, the signal distribution means 17 collects the data distribution and stores up to the state of the line 6. Next, at the time of detection, data is similarly accumulated up to the state of line f as shown in FIG. In FIG. 20B, since the object Ob is placed, a shadow is formed by the illumination from the illumination means 3a. Distribution comparison unit 3
20 in the normal state from the data of FIG.
20c is obtained by subtracting the data of (a) in FIG. The second object determination unit 33 detects this data as to whether it exceeds the threshold value T2 by the second threshold value setting unit 20, and determines the presence or absence of an object.

Normally, when there is no object, it is within the second threshold value ± T2, but when there is an object as shown in FIG. 20 (b), as shown in FIG. 2 There is a position that exceeds the threshold value ± T2 set by the threshold value setting unit 32. By doing this, an abnormal state can be found by comparison with the normal state.

An eleventh embodiment of the present invention will be described with reference to the drawings. In the eleventh embodiment, as shown in FIG. 21, 1 is an image input means for receiving the reflected light of the illumination means 3,
Reference numeral 4 is a lighting control means for controlling the lighting of the lighting means 3, 17 is a signal storage means for storing a differential signal between images when the lighting control means 4 controls different lighting states, and 34 is a signal storage means 17. A differential calculating means for calculating the position differential of the data from the accumulated data distribution, and 35 is an edge change detecting means for detecting the presence or absence of an object from the data of the differential calculating means 34.

The signal accumulating means 17 has a fifteenth structure, that is, an image memory 1 (19) and an image memory 2 (20).
And a difference calculation unit 21 that performs a difference calculation between the image memory 1 (19) and the image memory 2 (20), a storage memory 22 that stores the signal calculated by the difference calculation unit 21, and a storage memory 22.
The storage state determination unit 23 for determining the storage state and the storage control unit 24 for counting the number of storages or controlling the number of storages.

Then, the edge change detecting means 35 is
The peak detecting section 36 for detecting the peak value and its position from the data calculated by the differential calculating section 34, and the third threshold setting section 37 provided according to the number of times of accumulation of the data accumulated by the signal accumulating section 17 And a third object determination unit 38 that determines an object from the data detected by the peak detection unit 36 having a value equal to or higher than the third threshold value set by the third threshold value setting unit 37.

Next, the operation of the structure of the 11th embodiment will be described. FIG. 22 is a flow chart when the object recognition operation is started. An image of a state in which the illumination control means 4 turns off the illumination means 3 in step 101, and the image input means 1 turns off the illumination in the image memory 1 (19) of the signal storage means 17 in step 102 (hereinafter referred to as an off image). That)). Next, the state in which the lighting means 3 is turned on by the lighting control means 4 is set to step 10
In step 104, the image memory 2 (20) of the signal accumulating means 17 stores the image in the state in which the illumination is turned on (hereinafter referred to as an on-image) by the image forming means 1 created in step 3.

In step 105, the image memory 1 (19)
The difference calculation between the image memory 2 (20) and the image memory 2 (20) is performed by the difference calculation unit 21, and the calculation result is stored in the storage memory 22 in step 106. In step 107, the accumulation number control unit 11 counts the number of accumulations and checks whether accumulation is performed for a predetermined amount or more. If the accumulated amount is equal to or more than the predetermined amount, the process proceeds to step 109. If the number of accumulations is less than the predetermined number, the process proceeds to step 108. In step 108, the accumulation state determination unit 23 determines the peak value of the image data stored in the storage memory 22. If the peak value is less than or equal to the predetermined value, the process proceeds to step 101 and the image is input again by the image input means with the illumination turned off.

Here, the actual data will be described with reference to FIG. The graph of the accumulated data is a graph shown in FIG. 15 (a) when the object is not placed. In this graph, the horizontal axis represents position, and the vertical axis represents luminance difference (256 gradations). Illumination means 3a and 3b (marked with asterisks) are arranged on the left and right, and image input means 1 (⊚) is arranged in the center, so that they are overlapped on the graph. Since the illumination means 3a is placed at the left side position (marked with *), the left side shows a slightly higher value. Line 1, line 2, line 3, line 4, line 5, and line 6 are for storing image memory 1 (19) and image memory (7).
The image brightness difference is added up, and the state where the data accumulation is completed when the number of accumulation times is 6 is shown.

For example, the maximum value is 200 when the number of accumulations is set.
There are various methods such as a method of completing when the average value exceeds / 255, and a method of completing when the average value exceeds 180/255. Therefore, the number of times of accumulation may be determined in advance according to the brightness, or may be determined according to the above-described upper limit value of data. Similarly, FIG. 15B shows the case where there is no object.
And, the case where the illumination means 3b is placed on the right side is shown.

Once all the data has been accumulated, step 1
At 09, the data accumulated by the signal accumulating means 17 is subjected to the location differentiation by the differentiation calculating means 34. FIG. 23 shows data obtained by differentiating the distribution data of FIG. 15 by the differential calculating means 34. Figure 15
The data in (a), (b), (c), (d), (e), and (f) are shown in FIGS. 23 (a), (b), (c), (d),
It corresponds to (e) and (f).

Since the data in FIGS. 15A and 15B are data in which a smooth curve is drawn, the differential value is 0.
It is a value close to. On the other hand, since the data in FIGS. 15C and 15D have data changes at the end portions corresponding to the shadows, when the differentiation is performed, the differential value is a negative value when moving from the light side to the dark side, and from the dark side. When entering the brighter side, the differential value becomes a positive value, and the data has peaks as shown in (c) and (d) of FIG. In step 110, in order to detect an edge from the data, the data is input to the edge change detecting means 35, and the peak detector 36 detects the peak value and its position.

At step 111, the obtained peak value is the third value.
It is determined whether the value is greater than or equal to the threshold value T3 set by the threshold value setting unit 37. If it is greater than or equal to the threshold value T3, the third object determination unit 38 determines that there is an object in step 112, and if less than the threshold value T3, the third object determination unit 38 does not detect an object. To determine.

As described above, the existence of the object is recognized by detecting the edge.

The configurations described above are the fourteenth configuration, the case where the signal accumulating means 17 has the fifteenth configuration, and the case where the edge change detecting means has the sixteenth configuration. The edge change detection means 35 can determine the presence or absence of an object, which is the eighth object, if the same function is achieved by using another method.

In the eleventh structure, the lighting control means 4 turns the lighting means 3 on and off. However, since the purpose is to create a difference in illuminance, it is possible to extend the life of the lighting means and for other purposes. Therefore, it is possible to adopt a method of stopping the OFF state and illuminating in a weak state and creating a difference in illuminance difference depending on the intensity.

A twelfth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 24, in the twelfth embodiment, two lighting means 3a and 3b are provided as the lighting means 3 in the eleventh configuration and are installed at different positions. Also,
The second comparing means 39 for comparing the signals of the two edge change detecting means 35 and the second solid for judging whether or not the two data compared by the second comparing means 39 are different from each other and determining whether or not the object is a three-dimensional object. Object judging means 40 is added.

Next, the operation of the structure of the twelfth embodiment will be described.

In FIG. 23 (c), the object Ob is placed slightly to the right of the center. The illumination means 3a is the object O
It is designed to create a shadow of b. Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG. 23 (c), the object O
When there is b, the accumulated data changes the signal of the position of the end corresponding to the shaded portion.

On the other hand, in FIG. 23 (d), the illumination means 3a
Creates a shadow of the object Ob. Therefore, when there is an object Ob as shown in FIG. 23D, the signal of the position of the end portion corresponding to the shaded portion of the accumulated data changes due to the shadow.

The peak values of the left and right edges are detected as shown in the eighth embodiment, and the presence or absence of an object is determined. By the way, as can be seen from FIGS. 23 (c) and 23 (d), the positions of the edges are different due to the presence of a tall object due to the left and right illumination means 3a, 3b. Therefore, the position of the edge is compared by the second comparison means 39, and it is determined by the second three-dimensional object determination means 40 whether or not it is present at a different position. If it is at the different position, the object Ob is a three-dimensional object. judge.

Next, FIGS. 23 (e) and 23 (f) will be described. This is the case when the object has no height.

In FIG. 23 (e), the object Ob is placed slightly to the right of the center. The illumination means 3a is the object O
b, and the image input means 1 detects the difference in the brightness of the object. Now, assuming that the object Ob is a paper and has a large difference in brightness from the floor surface, the edge of the object is detected as an edge.

On the other hand, in FIG. 23 (f), the illumination means 3a
Illuminates the object Ob and inputs its reflection by the image input means 1. Now, assuming that the object Ob is a paper and has a large difference in brightness from the floor surface, the edge of the object is detected as an edge.
However, the positions of the detected edges are the same because the object has no height. Therefore, the position of the edge is compared by the second comparison means 39, and it is determined by the second three-dimensional object determination means 40 whether or not it is present at a different position. To do. As a matter of course, whether the positions are the same or not is considered to be the same position if a tolerance for left and right shifts is provided and the difference is within the tolerance.

As described above, it is possible to recognize a two-dimensional object and a three-dimensional object.

A thirteenth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 25, in the thirteenth embodiment, an opening / closing means 2, a drive section 29 for driving the opening / closing means 2, and a drive section control means 17 for controlling the drive section 29 are added to the eleventh configuration. .

Next, the operation of the structure of the thirteenth embodiment will be described. As described in the operation in the eleventh configuration, if the edge change detection means 35 has a peak value of edge data equal to or greater than the third threshold value T3, an object exists. The image pickup area of the image input means 1 is set to a portion where the end portion of the opening / closing means 2 corresponds to a closed state, and when an object exists at a position that hinders the opening / closing operation of the opening / closing means 2, the object determined previously is not crushed. The drive unit 29 is stopped via the drive control means 30. On the other hand, if the peak value of the edge data is less than the third threshold value T3 by the edge change detection means 35, there is no object. At that time, the drive unit 2 is driven through the drive control means 30.
9 is driven to operate the opening / closing means.

By doing so, safe operation can be performed even when an object is placed under the opening / closing means 2 such as a shutter.

A fourteenth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 26, the fourteenth embodiment has a twelfth structure in which an opening / closing means 2, a drive unit 29 for driving the opening / closing means 2, and a drive unit control means 17 for controlling the drive unit 29 are added. .

Next, the operation of the structure of the 14th embodiment will be described. As described in the operation in the twelfth configuration, the presence or absence of a three-dimensional object is detected by the second three-dimensional object determination means 40. Similar to the ninth embodiment, the object Ob is placed slightly to the right of the center in FIG. 23 (c). Lighting means 3
a creates a shadow of the object Ob. Since there is no reflection due to illumination in the shaded area, no light enters the image input means 1 (actually, attenuated light reflected from various places enters the image input means 1).

Therefore, as shown in FIG. 23 (c), the object O
When there is b, the accumulated data changes the signal of the position of the end corresponding to the shaded portion.

On the other hand, in FIG. 23 (d), the illumination means 3a
Creates a shadow of the object Ob. Therefore, when there is an object Ob as shown in FIG. 23D, the signal of the position of the end portion corresponding to the shaded portion of the accumulated data changes due to the shadow.

The peak values of the left and right edges are detected as shown in the eighth embodiment, and the presence or absence of an object is determined. By the way, as can be seen from FIGS. 23 (c) and 23 (d), the positions of the edges are different due to the presence of a tall object due to the left and right illumination means 3a, 3b. Therefore, the position of the edge is compared by the second comparison means 39, and it is determined by the second three-dimensional object determination means 40 whether or not it is present at a different position. If it is at the different position, the object Ob is a three-dimensional object. judge.

Next, FIG. 23 (e) and FIG. 23 (f) will be described. This is the case when the object has no height.

In FIG. 23 (e), the object Ob is placed slightly to the right of the center. The illumination means 3a is the object O
b, and the image input means 1 detects the difference in the brightness of the object. Now, assuming that the object Ob is a paper and has a large difference in brightness from the floor surface, the edge of the object is detected as an edge.

On the other hand, in FIG. 23 (f), the illumination means 3a
Illuminates the object Ob and inputs its reflection by the image input means 1. Now, assuming that the object Ob is a paper and has a large difference in brightness from the floor surface, the edge of the object is detected as an edge.
However, the positions of the detected edges are the same because the object has no height. Therefore, the position of the edge is compared by the second comparison means 39, and it is determined by the second three-dimensional object determination means 40 whether or not it is present at a different position. To do. As described above, it is possible to recognize a two-dimensional object and a three-dimensional object.

Now, when the image pickup area of the image input means 1 is set to a portion where the end portion of the opening / closing means 2 corresponds to the closed state, and a three-dimensional object exists at a position that obstructs the opening / closing operation of the opening / closing means 2, it is judged as previously. In order not to crush the three-dimensional object, the drive unit 29 is stopped via the drive control means 30. On the other hand, when the three-dimensional object determination means 15 does not exist, the drive unit 29 is driven via the drive control means 30 to operate the opening / closing means.

By doing so, safe operation can be performed even when a three-dimensional object is placed under the opening / closing means such as a shutter.

The operation of the fifteenth and sixteenth embodiments of the present invention has already been described in the portion explained in the eleventh configuration.

A seventeenth embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 27, in the seventeenth embodiment, the configuration of the edge change detecting means 35 is set to the differential calculating means 3 when normal.
Edge data calculated in 4 and differential calculation means 34 at the time of detection
The peak comparison unit 41 for comparing with the edge data calculated in
And a threshold value T for comparing the data of the peak comparison unit 41
4 is set by the fourth threshold value setting unit 42, and the fourth object determination unit 43 detects whether or not the data has a value of the threshold value T4 or more.

Next, the operation of the structure of the seventeenth embodiment will be described.

FIG. 28 is a diagram showing comparison data at the time of normal and at the time of detection. The vertical axis represents the brightness difference and the horizontal axis represents the position.

As shown in FIG. 28A, the edge data is sampled and stored by the differential calculating means 34 when it is considered to be normal in advance. Next, at the time of detection, as shown in FIG. 28B, the edge data is similarly sampled by the differential operation means 34. In FIG. 28B, since the object Ob is placed, a shadow is formed by the illumination from the illumination means 3a. The edge of this shadow is obtained as an edge, and the peak comparison unit 41 subtracts the differential data of FIG. 16A at the normal time from the differential data of FIG. 28B to obtain the data of FIG. 28C. To get This data is used by the fourth object determination unit 43 to determine the fourth
The threshold setting unit 30 detects whether or not the threshold T4 is exceeded, and determines the presence or absence of an object. Normally, if there is no object, it is within the fourth threshold value ± T4, but in FIG.
When there is an object as shown in FIG. 28, there is a position that exceeds the threshold value ± T4 set by the fourth threshold value setting unit 42 as shown in FIG. By doing this, an abnormal state can be found by comparison with the normal state.

Next, in the eighteenth embodiment of the present invention, the influence of sunlight and fluorescent lamp can be reduced by using the image input means 1 in the invisible region. The operation is as described above.

According to the above embodiments, the presence or absence of an object can be determined by its own illumination means without depending on the change in the environmental illuminance of the object to be recognized. Then, the fact that the data cannot be stored even by irradiation by the illumination means means that there is no reflection in that part, and the presence of the object or the limit of the capability of the image input means will be detected as a result. If the existence is recognized as abnormal, it will work on the safe side.

Further, it is possible to create different environments by controlling the illumination means at two places, and it is possible to recognize a three-dimensional object due to the difference in shadow. Then, the presence or absence of the three-dimensional object can be determined according to the degree of change in brightness near the three-dimensional object to be recognized.

By accumulating data for determining the presence / absence of an object, a large illuminance difference is obtained, and the accuracy of data distribution can be improved.

When the presence / absence of an object is determined, by changing the determination criterion according to the number of times of data accumulation, the noise component increases and the noise component increases. Therefore, the influence can be removed, and the malfunction can be determined less. .

It is possible to make a judgment with few malfunctions by using a predetermined judgment criterion based on a difference from normal data for judging the presence or absence of an object.

The presence or absence of an object can be determined by its own illumination means without depending on the change in the environmental illuminance of the object to be recognized. Then, the edge feature of the object can be found in the determination of the presence or absence of the object, and the determination can be made with less malfunction. When an object is placed on a uniform floor surface, the space between the object and the floor surface is detected as an edge, and the presence / absence of the object is determined more easily than comparing the distribution of raw data.

Different environments can be created by controlling the illumination means at two places, and the three-dimensional object can be recognized due to the difference in shadow. Then, the end feature of the three-dimensional object can be found in the determination of the presence or absence of the three-dimensional object, and the determination can be made with less malfunction.

By accumulating data for determining the presence / absence of an object, a large illuminance difference is obtained, and the accuracy of edge detection can be improved.

The edge feature of the object can be determined with less malfunction according to the degree of the peak value of the edge.

It is possible to make a judgment with few malfunctions by using the judgment criterion based on the difference between the edge feature of the object and the data at the normal time.

By detecting the object in the invisible region, it is possible to prevent the person from noticing it. By using the infrared LED as the illumination in the invisible region, the response is faster than that of a halogen lamp often used in the visible illumination. To secure the accumulated data, it takes a long time to repeatedly sample the data, but since the rise and fall times at the time of on / off are short, the total time can be greatly shortened. It should be noted that if a visible LED is used when it is visible, the entire environment changes due to its color, which may be inappropriate depending on the installation location. However, it does not affect the environment when illuminated in the invisible region.

[0182]

As described above, according to the present invention, the presence or absence of an object can be determined by its own illumination means without depending on the change in the environmental illuminance of the object to be recognized. And the fact that the data cannot be stored even by irradiation by the illumination means that there is no reflection in that part,
When the object exists or the limit of the ability of the image input means is detected as a result, when the existence of the object is recognized as abnormal, it works on the safety side.

[Brief description of drawings]

FIG. 1 is a configuration block diagram of an object recognition device of an automatic opening / closing device according to a first embodiment of the present invention.

FIG. 2 is a configuration block diagram of an object recognition device of an automatic opening / closing device according to a second embodiment of the present invention.

FIG. 3 is a diagram showing an image input by an object recognition device of an automatic opening / closing device.

FIG. 4 is a diagram showing an image input by an object recognition device of an automatic opening / closing device.

FIG. 5 is a diagram showing an image input by an object recognition device of an automatic opening / closing device.

FIG. 6 is a diagram showing an image input by an object recognition device of an automatic opening / closing device.

FIG. 7 is a diagram showing image processing data in an object recognition device of an automatic opening / closing device.

FIG. 8 is a diagram showing image processing data in an object recognition device of an automatic opening / closing device.

FIG. 9 is a diagram showing image processing data in an object recognition device of an automatic opening / closing device.

FIG. 10 is a diagram showing image processing data in the object recognition device of the automatic opening / closing device.

FIG. 11 is a diagram showing image processing data in the object recognition device of the automatic opening / closing device.

FIG. 12 is a diagram showing image processing data in the object recognition device of the automatic opening / closing device.

FIG. 13 is a block diagram of an object recognition device according to a fourth embodiment of the present invention.

FIG. 14 is a flowchart of the object recognition device.

FIG. 15 is a diagram showing image data input by the object recognition device.

FIG. 16 is a block diagram of an object recognition device according to a fifth embodiment of the present invention.

FIG. 17 is a block diagram of an object recognition device according to a sixth embodiment of the present invention.

FIG. 18 is a block diagram of an object recognition device according to a seventh embodiment of the present invention.

FIG. 19 is a block diagram of an object recognition device according to a tenth embodiment of the present invention.

FIG. 20 is a diagram showing image data input by the object recognition device.

FIG. 21 is a block diagram of an object recognition device according to an eleventh embodiment of the present invention.

FIG. 22 is a flowchart of the object recognition device.

FIG. 23 is a diagram showing image data input by the object recognition device.

FIG. 24 is a block diagram of an object recognition device according to a twelfth embodiment of the present invention.

FIG. 25 is a block diagram of an object recognition device according to a thirteenth embodiment of the present invention.

FIG. 26 is a block diagram of an object recognition device according to a fourteenth embodiment of the present invention.

FIG. 27 is a block diagram of an object recognition device according to a seventeenth embodiment of the present invention.

FIG. 28 is a diagram showing an example of image data input by the object recognition device.

FIG. 29 is a block diagram of a conventional automatic opening / closing device.

[Explanation of symbols] 1 Image input means 2 opening and closing means 3 lighting means 4 Lighting control means 5 Illuminance correction means 6 Object recognition means 7 Drive unit control means 8 Image storage means 9 Illuminance measuring means 10 Image difference means 11 Means of determining the number of accumulations 12 Difference image storage means 16 Specific background means 17 Signal storage means 18 Distribution change detection means 19 Image memory 1 20 image memory 2 21 Difference calculator 22 Storage memory 23 Accumulation state determination unit 24 Storage control unit 25 First Threshold Setting Unit 26 First Object Determination Unit 27 First Comparison Means 28 First three-dimensional object determination means 29 Drive 30 Drive control means 31 Distribution comparison section 32 Second threshold setting unit 33 second object determination unit 34 Differential calculation means 35 Edge Change Detection Means 36 Peak detector 37 Third Threshold Setting Unit 38 Third Object Judgment Unit 39 Second comparison means 40 Second three-dimensional object determination means 41 Peak comparison section 42 Fourth Threshold Setting Unit 43 Fourth Object Judgment Unit

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Shinji Miyauchi             1006 Kadoma, Kadoma-shi, Osaka Matsushita Electric             Sangyo Co., Ltd. F-term (reference) 2E052 AA01 AA04 BA06 GA06 GB01                       KA01                 5B057 AA19 BA02 BA19 BA30 DA07                       DB02 DB09 DC05 DC16 DC22                       DC32                 5L096 AA06 BA02 CA04 EA12 FA06                       FA14 FA35 GA02 GA08 GA51

Claims (11)

[Claims] 1. An illumination unit, an image input unit for receiving reflected light of the illumination unit, an illumination control unit for controlling illumination of the illumination unit, and a case where the illumination control unit controls different illumination states. 2. An object recognition apparatus having a signal accumulating means for accumulating a differential signal between the images and a distribution change detecting means for detecting the presence or absence of an object from the data distribution accumulated by the signal accumulating means. 2. The first comparing means for comparing two data obtained by the distribution change detecting means obtained when the two illuminating means at different positions are individually turned on, and the first comparing means. The object recognition device according to claim 4, further comprising a first three-dimensional object determination unit that determines that a three-dimensional object is detected by detecting that the compared two data are different. 3. The signal storage means includes an image memory 1, an image memory 2, a difference calculation section for calculating a difference between the image memory 1 and the image memory 2, and a storage memory for storing signals calculated by the difference calculation section. 2. A storage state determination unit that determines a signal level stored in the storage memory, and a storage control unit that controls the number of times storage is performed in the storage memory.
The object recognition device described. 4. The distribution change detection means is provided with a first threshold value setting section provided according to the number of times of accumulation of the data accumulated by the signal accumulating means, and data having a value equal to or greater than the first threshold value set by the first threshold value setting section. A first object determination unit for detecting
The object recognition device described. 5. The distribution change detecting means is provided by the distribution comparing means for comparing the data distribution accumulated by the signal accumulating means at the normal time with the data distribution accumulated by the signal accumulating means at the time of detection, and the distribution comparing means. The object recognition device according to claim 1, further comprising a second object determination unit that detects data having a value equal to or greater than a second threshold value. 6. An illumination means, an image input means for receiving the reflected light of the illumination means, an illumination control means for controlling the illumination of the illumination means, and when the illumination control means controls different illumination states. Signal accumulating means for accumulating a differential signal between the images, differential calculating means for calculating a differential value of the data distribution accumulated by the signal accumulating means, and presence / absence of an object are detected from the data calculated by the differential calculating means. An object recognition device having edge change detection means. 7. A second comparing means for comparing the two data obtained by the distribution differentiation calculating means obtained when the two illuminating means at different positions are individually turned on, and the second comparing means. 7. The object recognition device according to claim 6, further comprising a second three-dimensional object determining unit that detects a three-dimensional object by detecting that the compared two data are different. 8. The signal storage means includes an image memory 1, an image memory 2, a difference calculation section for calculating a difference between the image memory 1 and the image memory 2, and a storage memory for storing the signals calculated by the difference calculation section. 7. A storage state determination unit that determines a signal level stored in the storage memory, and a storage control unit that controls the number of times storage is performed in the storage memory.
The object recognition device described. 9. The edge change detecting means detects a peak of the data calculated by the differential calculating means, and a peak detecting section for detecting data having a peak value of not less than a third threshold value detected by the peak detecting section. 7. A three-object determination unit
The object recognition device described. 10. An edge change detecting means is provided in the peak comparing section, which compares the data calculated by the differential calculating means during normal operation with the data calculated by the differential calculating means during detection, and the peak comparing section. The object recognition device according to claim 6, comprising a fourth object determination unit that detects data having a value equal to or greater than a threshold value. 11. The object recognition device according to claim 1, wherein the illumination means and the image input means are used in an invisible area.