JP2012155917A - Lighting control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a lighting control system that can both enhance the detection accuracy of the existence or non-existence of a human and increase the processing speed of detecting the existence or non-existence of a human.SOLUTION: A lighting control system 10 comprises a human detecting sensor unit 5 having an imaging device 1 and an image processor 2 for detecting the existence of a human by executing image processing on an image obtained by the imaging device 1, and a controller 6 for controlling illumination light on the basis of a detection signal from the human detecting sensor unit 5. The image processor 2 executes first image processing for detecting the existence of a human in a predetermined overall range of the image, and second image processing for dividing the predetermined overall range into plural different specific ranges, counting the frequency of the detection of the existence of a human during a predetermined time for every specific range, and detecting the existence of a human for the specific ranges which are narrower than the predetermined overall range and whose counted frequencies exceed a predetermined frequency.

Description

  The present invention relates to a lighting control system.

  2. Description of the Related Art Conventionally, there has been developed an illumination control system that controls a light source unit provided in a lighting fixture by detecting the presence or absence of a person in an image captured and generated using an imaging device. The lighting control system can be used for crime prevention by turning on the light source unit when the presence of a person is detected, or can contribute to energy saving by turning off the light source unit when the presence of a person cannot be detected.

  As this type of lighting control system, for example, as shown in FIG. 9, a TV camera provided with an imaging device such as a CCD image sensor (Charge Coupled Device Image Sensor) that images a room 30 that is subject to lighting control in an office or the like. 31, an image obtained from an image signal from the TV camera 31 is divided into a plurality of areas, a brightness is detected for each of the divided areas, and the presence or absence of moving objects is detected. What is provided with the lighting fixtures 7 and 7 by which light output is controlled by control is known (for example, patent document 1).

  The illumination control system of Patent Document 1 described above detects the presence or absence of a moving object from changes in images continuously obtained from the TV camera 31, and detects the presence or absence of a moving object as the presence or absence of a person. Further, the lighting control system calculates the brightness in the brightness detection area LDA by the control device 32 based on the image signal from the TV camera 31 and controls the light output of the lighting fixtures 7 and 7 based on the result. The signal Sd is output. The illumination control system is capable of performing illumination control for turning off the corresponding lighting fixtures 7 and 7 for each area when there is no person.

  In addition, as a monitoring camera device attached to an outer wall of a house, for example, an imaging camera having an imaging device such as a CCD image sensor and a grayscale image of image data captured and generated by the imaging camera are subjected to differentiation processing. An image processing unit that recognizes a moving object by performing image processing for extracting a contour image and detects a person from the movement of the moving object, and a signal to the illuminating lamp to irradiate the detected person with an illuminating lamp as an intruder A device including a signal processing unit for outputting is also known (for example, Patent Document 2).

JP 2002-289377 A JP 2007-318674 A

  By the way, in the lighting control system, in order to detect the presence / absence of a person with higher accuracy, the above-described monitoring camera device is used for image data generated by imaging the TV camera 31 of the above-described lighting control system. It is conceivable to perform image processing that extracts a specific contour image as it is performed.

  In this case, in order for the lighting control system to accurately detect the presence or absence of a person, the lighting control system needs to perform image processing on, for example, about 30 images per second obtained by the imaging device. is there. The lighting control system often has a wide range for detecting the presence or absence of a person, and it may be preferable to detect the presence or absence of a stationary person as well as a moving person. In order for the lighting control system to detect the presence / absence of a person at high speed and with high accuracy, for example, a shape pattern recognition process is performed on the result of feature extraction such as a contour image for image data generated by imaging. It is also necessary to perform image processing for extracting a person by performing image identification processing such as.

  However, in the illumination control system, in the case of image processing in which image identification is performed on the result of feature extraction, the amount of data processing is large, and the load on the image processing apparatus that performs image processing becomes very large. In addition, the number of pixels of image pickup devices has been increasing year by year, and although the resolution can be increased, the number of pixels of image data in an image per image obtained by the image pickup device tends to increase significantly. is there. In the illumination control system, the amount of data processing consumed for image processing of image data per sheet is increasing, and the performance of image processing required for the image processing apparatus is higher. For this reason, the illumination control system may increase the cost of the entire illumination control system including the image processing apparatus. That is, in the lighting control system, there is a trade-off relationship between improvement in detection accuracy of the presence / absence of a person and processing speed for detecting the presence / absence of a person.

  The present invention has been made in view of the above reasons, and its purpose is to improve the detection accuracy of the presence / absence of a person and the processing speed to detect the presence / absence of a person with a relatively simple configuration. It is to provide a lighting control system that can achieve both.

  An illumination control system of the present invention includes an imaging device and an image processing unit that detects the presence of a person by performing image processing on an image based on image data generated by imaging a specific area by the imaging device and outputs a detection signal And a control unit that controls illumination light emitted from the light source unit based on a detection signal from the human detection sensor unit, wherein the image processing unit includes: As the image processing, a first image processing for detecting the presence of a person by performing image identification processing on a result of feature extraction performed on the entire predetermined range of the image, and a plurality of the entire predetermined range. The number of times the presence of a person is detected for each specific range divided into a specific range is counted during a predetermined time, and after the predetermined time has elapsed, the number is narrower than the entire predetermined range. And the counted frequency and performing a second image processing for detecting the presence of human with respect to the specific range exceeds a predetermined frequency.

  In the illumination control system, when the human detection sensor unit detects a specific marker set in advance by performing image identification processing within the predetermined range, the human detection sensor unit detects an area where the marker is detected after the detection. It is preferable that the specific range in which the image identification process for detecting the presence of the person is performed, or the region where the marker is detected is outside the specific range in which the image identification process for detecting the presence of the person is detected.

  In the illumination control system, the control unit is connected to a plurality of the light source units, and the control range of the illumination light emitted from the light source unit is the specific range for performing image identification processing for detecting the presence of a person. It is preferable that the difference is outside the specific range.

  In the illumination control system, the control content is preferably the brightness of illumination light emitted from each of the plurality of light source units.

  In this illumination control system, the content of the control is the illumination emitted from the light source unit after the control unit starts controlling the illumination light emitted from the plurality of light source units to have a predetermined brightness. It is preferable that the time is for each light to reach the predetermined brightness.

  In the illumination control system, the control content includes brightness of illumination light emitted from each of the plurality of light source units, and control of the illumination light emitted from the light source unit to a predetermined brightness by the control unit. It is preferable that the period of time from the start of the illumination until the illumination light respectively emitted from the light source unit reaches the predetermined brightness.

  The illumination control system of the present invention has an effect that it is possible to achieve both improvement in detection accuracy of presence / absence of a person and improvement in processing speed for detecting presence / absence of a person with a relatively simple configuration.

1 is a schematic configuration block diagram illustrating an illumination control system of Embodiment 1. FIG. The lighting control system same as the above is shown, (a) is a diagram for explaining the arrangement of components, (b) is a reference diagram for an image generated by imaging, and (c) is a diagram for explaining the principle of image processing. It is a flowchart figure which shows operation | movement of a lighting control system same as the above. It is explanatory drawing of the image process in an illumination control system same as the above. It is explanatory drawing of another image processing in an illumination control system same as the above. It is explanatory drawing of the image processing in the illumination control system of Embodiment 2. The lighting control system of Embodiment 3 is shown, (a) is a schematic block diagram, (b) is a principal part block diagram. It is explanatory drawing of the image process in an illumination control system same as the above. It is a block diagram of the conventional illumination control system.

(Embodiment 1)
Hereinafter, the illumination control system 10 of the present embodiment will be described with reference to FIGS. 1 to 5.

  As shown in FIG. 1, the illumination control system 10 of the present embodiment applies image data generated by imaging the imaging device 1 and the imaging device 1 by imaging a specific area S1 (see the broken line in FIG. 2A). A human detection sensor unit 5 including an image processing unit 2 that detects the presence of the person H by performing image processing on an image based on the image (see FIG. 2B) and outputs a detection signal is provided. Further, the illumination control system 10 includes a control unit 6 that controls illumination light emitted from the light source unit 7 a provided in the lighting fixture 7 based on a detection signal from the human detection sensor unit 5.

  In the illumination control system 10, the image processing unit 2 performs image identification processing based on shape pattern recognition on the result of feature extraction of a contour image such as a person H in the entire predetermined range of the image as the image processing. By performing the first image processing, the presence of the person H is detected. In addition, the lighting control system 10 divides the entire predetermined range into a plurality of different specific ranges, counts the number of times the presence of the person H is detected for each specific range, and determines the predetermined range. After the elapse of time, the second image processing for detecting the presence of the person H is performed on the specific range that is narrower than the entire predetermined range and the counted frequency exceeds the predetermined frequency. . Here, the predetermined range may be the entire image, or may be a range in which an area that does not require image processing in advance, such as a column in a room, is excluded from image processing targets.

  More specifically, in the illumination control system 10 of the present embodiment, for example, analog data corresponding to about 30 images per second obtained by the imaging device 1 provided in the human detection sensor unit 5 is digitized. The data is output as needed to the calculation control unit 3 in the subsequent image processing unit 2 in units of pixels. As such an imaging device 1, for example, a solid-state imaging device such as a CCD image sensor (Charge Coupled Device Image Sensor) or a CMOS image sensor (Complementary Metal Oxide Semiconductor Image Sensor) can be used.

  The image processing unit 2 includes, for example, an arithmetic control unit 3 that performs arithmetic processing on image data obtained by digitizing analog data that can be directly acquired from the imaging device 1, and a first storage unit 4a that stores image data from the imaging device 1. The second storage unit 4b stores in advance data on the shape of the contour for performing image identification processing.

  The arithmetic control unit 3 can be configured using a microcomputer or the like. The arithmetic control unit 3 is a semiconductor capable of processing digital image processing such as a DSP (Digital Signal Processor) and an advanced image processor at high speed in order to perform image processing of image data sequentially transferred from the imaging device 1 with quick response. An element can be provided.

  In addition, since the image processing unit 2 has a large amount of data corresponding to the image obtained by the imaging device 1, the image processing unit 2 uses a mass storage device that stores the image data from the imaging device 1 in the image processing unit 2. One storage unit 4a is preferably provided. In order to perform image identification processing, the image processing unit 2 includes a second storage unit 4b for storing contour shape data such as a person H in advance separately from the first storage unit 4a. That is, the human detection sensor unit 5 is configured to temporarily store the image data from the imaging device 1 in the first storage unit 4a and appropriately read the image data into the arithmetic control unit 3 and use it for image processing. Similarly, the human detection sensor unit 5 is configured such that data of a contour shape such as a person H can be appropriately read from the second storage unit 4b into the arithmetic control unit 3 and used for image processing.

  As the first storage unit 4a for storing image data, for example, a storage device including a volatile memory having a relatively large storage capacity such as a DRAM (Dynamic Random Access Memory) can be used. In addition, the second storage unit 4b for preliminarily storing contour shape data and the like for performing image identification processing includes a non-volatile memory capable of transferring image data at a comparatively high speed, such as SDRAM (Synchronous DRAM). It can also be used.

  The human sensor unit 5 may be configured by integrating the imaging device 1 and the image processing unit 2 or may be configured by physically separating them. The number of times the image processing unit 2 detects the presence of the person H for each specific range may be stored in either the first storage unit 4a or the second storage unit 4b.

  Here, as shown in FIG. 2A, the illumination control system 10 of the present embodiment includes, for example, a person detection sensor unit 5 that detects the presence or absence of a person H on the ceiling surface CS side, and the person detection. A luminaire 7 electrically connected to the sensor unit 5 is disposed. In the illumination control system 10, the imaging device 1 of the human detection sensor unit 5 images a specific area S1 from the ceiling surface CS side, detects the presence or absence of a person H in the specific area S1, and detects the human detection sensor unit 5 Is configured to automatically turn on the lighting fixture 7. In the illumination control system 10, the imaging device 1 of the human detection sensor unit 5 images the vertically downward side from the ceiling surface CS side. Therefore, in the image based on the image data generated by being imaged by the imaging device 1, for example, the person H appears as shown in FIG. Here, the human detection sensor unit 5 separately stores in advance the image data before the person H enters the specific area S1 in the first storage unit 4a as background data. The human detection sensor unit 5 outputs image data generated from data output from the imaging device 1 and background data stored in the first storage unit 4a from the imaging device 1 side. The person H who entered the specific area S1 is detected by comparing each time.

  That is, as shown in FIG. 2C, the human detection sensor unit 5 is configured so that the image processing unit 2 has pixel values for each pixel of the current image C1 based on the current image data and the background image C2 based on the background data. The difference image C3 which took the difference is generated. Note that the current image C1 and the background image C2 each represent a pixel value of a pixel as a luminance value. Further, the difference image C3 represents the luminance value in the pixel value of each pixel in a binarized manner. At this time, the current image C1 and the background image C2 are compared, and the luminance value of the pixels that do not change between the current image C1 and the background image C2 is 0 by taking the difference. Therefore, in the binarized difference image C3, the above-described pixels that do not change can be expressed as white, for example, below a predetermined threshold (see the white square of the difference pixel C3). On the other hand, when the current image C1 and the background image C2 are compared and a person H enters, a pixel that changes between the current image C1 and the background image C2 (hereinafter referred to as a difference pixel) is binary. As a result, the predetermined threshold value can be expressed as black, for example (see the black square of the difference image C3).

  In the calculation control unit 3 of the image processing unit 2, the total number of difference pixels in the difference image C3 generated by comparing the current image C1 and the background image C2, and the size of a pixel group in which consecutive or adjacent difference pixels are gathered. Etc. can be detected.

  Next, the operation in the illumination control system 10 of the present embodiment will be described in detail based on the flowchart of FIG.

  As shown in FIG. 3, in the illumination control system 10, as the image processing for detecting the presence / absence of the person H, in the image processing unit 2 of the human detection sensor unit 5, first, the calculation control unit 3 has the first Image difference processing A is performed in which a difference between pixel values for each pixel of the two image data (current image C1 and background image C2) stored in the storage unit 4a is obtained. The human detection sensor unit 5 stores image data corresponding to at least two images (for example, the current image C1 and the background image C2) in the first storage unit 4a for the image difference processing A of the image data. Yes. The arithmetic control unit 3 may appropriately read the image data generated by the imaging device 1 and perform image difference processing A.

  Further, the two image data used for the image difference processing A are not limited to the current image C1 and the background image C2 based on the background data before the person H enters the specific area S1. For example, as two pieces of image data used for the image difference processing A, adjacent image data in time series generated by continuously imaging the imaging device 1 can be used. That is, the human detection sensor unit 5 may update the current image C1 one by one as the background image C2 in time series.

  Next, the image processing unit 2 performs a difference pixel extraction process B for extracting the difference pixels in the two image data acquired by the image difference process A so that the difference image C3 can be generated.

  If there is a change between two pieces of image data due to the movement of a moving object such as a person H, the human detection sensor unit 5 is detected as a pixel group in which the difference pixels are gathered to a certain size. When the human detection sensor unit 5 extracts a pixel group that is a set of the above difference pixels that is equal to or larger than a predetermined size set in advance by the image processing unit 2 by the difference pixel extraction process B, the person H who is moving is detected. A detection process F of the person H (motion) who is determined to be present is performed (see the lower part of FIG. 3).

  For example, the human detection sensor unit 5 compares the number of pixels detected by the arithmetic control unit 3 with a predetermined threshold value, and if the detected pixel number exceeds the threshold value, the human detection sensor unit 5 falls within the predetermined range. It may be determined that H has entered. The human detection sensor unit 5 detects the intrusion of the human H into the predetermined range by the human H (motion) detection process F, and turns on the light source unit 7 a included in the lighting fixture 7. The arithmetic control unit 3 outputs a detection signal to the control unit 6.

  The control unit 6 that has received the detection signal from the human detection sensor unit 5 side performs, for example, a lighting control process G that controls the power supply to the light source unit 7a provided in the lighting fixture 7 and controls the lighting of the light source unit 7a. .

  Based on the detection signal from the human detection sensor unit 5 side, the control unit 6 starts to supply power to the light source unit 7 a included in the lighting fixture 7 and turns it on. The control unit 6 outputs a control signal for dimming the light source unit 7a to a predetermined brightness based on a detection signal from the human detection sensor unit 5, and the illumination light irradiated from the light source unit 7a is set to a predetermined brightness. Also good.

  For example, the control unit 6 includes a timer unit (not shown), and starts to count down the timer unit in synchronization with receiving a detection signal from the human detection sensor unit 5. The control unit 6 lights the light source unit 7a included in the lighting fixture 7 during the lighting holding time until the timer unit completes the countdown. The control unit 6 controls to turn off the light source unit 7a when the timer unit completes the countdown and the lighting holding time elapses without receiving a new detection signal from the human detection sensor unit 5. When the timer unit receives a new detection signal from the human detection sensor unit 5 while the timer unit is counting down, the control unit 6 resets the count of the timer unit. The control unit 6 keeps the light source unit 7a lit during the lighting holding time until the timer unit resets and completes a new countdown. That is, the lighting time of the light source unit 7a is extended by the amount that the count of the timer unit is reset. The timer unit may not only count down, but may measure the lighting maintenance time by counting up until a predetermined count is reached.

  The control part 6 should just be what can control the illumination light irradiated from the light source part 7a based on the detection signal from the human sensitive sensor part 5. FIG. Moreover, as the light source part 7a, for example, a discharge lamp such as a fluorescent lamp, a mercury lamp, an incandescent light bulb, a light emitting diode, or the like can be used. The control unit 6 may be appropriately set according to the type of the light source unit 7a. For example, when a fluorescent lamp is used as the light source unit 7a, an electronic ballast using an inverter circuit may be used. In addition, when the light emitting diode is used as the light source unit 7a, the control unit 6 is an AC-DC converter that converts commercial AC power supplied from the outside into DC power, and a drive control that controls current limiting resistance and current flowing through the light emitting diode. What is necessary is just to comprise suitably by a part.

  The luminaire 7 may include the light source unit 7a, may be formed integrally with the light source unit 7a, or may be configured to replace the light source unit 7a. The lighting fixture 7 may or may not include the control unit 6.

  Here, when the human detection sensor unit 5 always uses the same background image C2 to perform the differential pixel extraction process B, the illumination control system 10 is configured so that, for example, a stationary object other than the person H is the predetermined object. When entering the range, the lighting state of the lighting fixture 7 may be continued due to the presence of a stationary object other than the person H detected by the person detection sensor unit 5.

  For this reason, the illumination control system 10 is configured to use the human detection sensor unit 5 that updates the current image C1 in time series as the background image C2 one by one, and the human detection sensor that always uses the same background image C2. Since the stationary object other than the person H can be backgrounded than the case where the unit 5 is used, the presence of the person H can be detected with higher accuracy in many cases.

  However, in the illumination control system 10 configured using the human detection sensor unit 5 that updates the current image C1 one by one as the background image C2 in time series, when the human H is stationary, the human detection sensor unit 5 uses the human H as a background. It will be processed. Therefore, in the illumination control system 10, it is difficult for the human detection sensor unit 5 to discriminate between the person H and a stationary object other than the person H.

  That is, the lighting control system 10 includes the human detection sensor unit 5 that always uses the same background data and the human detection sensor unit 5 that only updates the background data in time series, so that the person H operates. It is considered effective for detecting the presence or absence of the person H in the state where the person is present. However, it is not sufficient for the illumination control system 10 to include the above-described human detection sensor unit 5 in order to detect the person H whose human detection sensor unit 5 is stationary with higher accuracy.

  Therefore, the illumination control system 10 of this embodiment includes a human detection sensor unit 5 that performs the following processing.

  That is, the human detection sensor unit 5 performs the recognition process C for the difference pixel after performing the extraction process B of the difference pixel (see the upper stage in FIG. 3). In the recognition process C, it is determined how much the result of the feature extraction obtained by extracting the contour shape of the group of difference pixels matches the data of the predetermined contour shape stored in advance in the second storage unit 4b. Image identification processing by shape pattern recognition is performed. The second storage unit 4b stores in advance a pattern of a person H or an object shape when viewed from the ceiling surface CS side. The pattern of the shape of the person H is not limited to one type, and a plurality of types of patterns may be stored. The shape pattern recognition process refers to a process for recognizing the coincidence rate with the shape of the person H stored in advance in the pixel group of the difference pixels.

  Since the change of the luminance value of the pixel is large at the boundary that becomes the outline of the group of difference pixels, if the differential calculation or the difference calculation is performed on the change of the luminance value, the shape of the outline of the group of the difference image is changed. Can be detected. For example, the human detection sensor unit 5 can extract the shape of the contour using a Sobel filter, a Prewit filter, or the like for the density of the binarized image based on the image data output from the imaging device 1.

  The arithmetic control unit 3 determines whether the extracted contour shape is due to the person H by determining the matching rate between the extracted contour shape and the pattern stored in the contour shape of the person H stored in advance. It is possible to determine whether it does not fall under H. Therefore, in the recognition process C for the difference pixel, the human detection sensor unit 5 does not detect the person H even if the person H is in a stationary state where the person H can be regarded as substantially not moving. The person H (stationary) detection process D can be performed to determine whether or not there is any.

  As a result, the illumination control system 10 including the imaging device 1 detects the presence or absence of the person H and controls the illumination light even when the person H is in a stationary state. Can be done.

  It should be noted that the extracted contour shape and the pre-stored pattern in the contour shape of the person H can determine the matching rate by template matching or the like.

  When the human detection sensor unit 5 determines that the human H is present by the detection process D for the human H (stationary), the calculation control unit 3 of the image processing unit 2 outputs the detection signal output by the lighting control process G and Outputs different predetermined detection signals. When the control unit 6 receives the predetermined detection signal, unlike the lighting control process G described above, the control unit 6 does not count down the timer unit for lighting the light source unit 7a during the lighting holding time. The illumination control system 10 continues the detection process D of the person H (stationary), and turns on the light source part 7a provided in the lighting fixture 7 until the predetermined detection signal is not output from the person detection sensor part 5 side. Continued lighting control processing E is performed. The human detection sensor unit 5 detects the predetermined detection signal if the presence of the human H is not detected by the movement of the human H from the predetermined range while the human H (stationary) detection process D is continued. Will not be output.

  Next, the first image processing and the second image processing will be described in detail as image processing in the illumination control system 10 of the present embodiment.

  In the lighting control system 10, as the first image processing, during the lighting control processing E, image processing is performed on the result of the feature extraction of the contour image such as the person H in the entire predetermined range of the image. The unit 2 performs image identification processing based on shape pattern recognition.

  For example, the number of times the image processing unit 2 determines the person H for each specific range by dividing the entire predetermined range of the image into a plurality of different specific ranges in a lattice shape for the second image processing. Is stored in the second storage unit 4b. The second storage unit 4b is provided with a storage area corresponding to each specific range in a lattice shape. When the image processing unit 2 detects a specific range determined as the person H during a predetermined time, the count value in the storage area corresponding to the specific range is incremented by one each time the person H is detected. And count the number of times the presence of the person H is detected.

  Specifically, as a plurality of different specific ranges obtained by dividing the entire predetermined range of the image in a grid pattern, for example, an image based on image data generated by the imaging device 1 imaging a specific region S1 is displayed on the vertical axis. FIG. 4 illustrates an 8 × 13 divided area (hereinafter referred to as a captured image) having a to h (eight) areas in the direction and 0 to 12 (13) areas in the horizontal axis direction. Hereinafter, the position of each specific range is specified by using a to h in the vertical axis direction and 0 to 12 in the horizontal axis direction, for example, the upper left corner of the captured image in FIG.

  The captured image shown in FIG. 4 includes, for example, an object such as a desk recognized in the background image C2 (four desks 21 in FIG. 4) and a pixel group regarded as a person H by performing image identification processing. A plurality of (here, four) specific ranges R2 to which P belongs (see the regions f7, f8, g8, and g7 in FIG. 5) are illustrated.

  As the second image processing, the human detection sensor unit 5 divides the entire captured image, which is a predetermined range, into 8 × 13 specific ranges, and determines the number of times the presence of the person H is detected for each specific range. Counting is performed during a predetermined time during the lighting control process E. In the second storage unit 4b in the human detection sensor unit 5, the count value for each specific range in the image data changes in the storage area. After a predetermined time has elapsed, the image processing unit 2 compares the count value stored in the second storage unit 4b with a predetermined frequency set in advance as a threshold value. As a result of the comparison, the image processing unit 2 can detect the specific range R1 having a count narrower than the entire predetermined range and lower than the predetermined frequency, and the specific range R2 having a predetermined frequency or more. After a predetermined time, the image processing unit 2 detects the presence of the person H by performing image identification on the result of the feature extraction performed in the specific range R2 having a predetermined frequency or more. In other words, the image processing unit 2 does not perform image identification on the result of feature extraction in a specific range R1 lower than a predetermined frequency after a predetermined time.

  That is, as shown in FIG. 5, the image processing unit 2 has a specific range R2 (regions a2, a3, a4, b4, b3 and b2, and a6, a7, a8, b8, b7 and b6 in FIG. 5). Area, e3, e4, f4, g4, g3 and 3f areas, e7, e8, f8, g8, g7 and f7 areas, areas from b10 to g10 and b11 to f11 in the vertical axis direction) It is defined as an area for processing. In other words, the lighting control system 10 learns the frequency of the stationary person H until a predetermined time elapses in order to perform the second image processing for detecting the presence of the person H for the specific range R2. By doing so, the range for automatically identifying the image is set to be narrowed.

  As the image processing after a predetermined time has elapsed, the image processing unit 2 performs the image difference processing A and the difference pixel extraction processing B as the person H (motion) detection processing F over the entire predetermined range in the captured image. However, for the detection process D of the person H (stillness) after the recognition process C for the difference pixel, the image processing unit 2 detects a specific number in which the number of times the presence of the person H is detected during a predetermined time exceeds a predetermined frequency. Control is performed only for the range R2.

  As a result, the illumination control system 10 can significantly reduce the processing load on the image processing unit 2 as compared with the one that performs the image identification process over the entire predetermined range of the image based on the image data. In addition, it is possible to suppress the detection omission of the stationary person H and erroneous detection of a stationary object other than the person H as the person H, and to detect the presence or absence of the person H with higher accuracy.

(Embodiment 2)
The lighting control system 10 of the present embodiment sets the specific range R2 for performing the second image processing of the first embodiment shown in FIG. 5 to be automatically narrowed from the predetermined range after the predetermined time has elapsed. In addition to this, the specific range R2 in which the second image processing is performed can be changed by detecting a specific marker M previously set at a predetermined position in the captured image shown in FIG. In addition, the same code | symbol is attached | subjected to the component similar to Embodiment 1, and description is abbreviate | omitted suitably.

  The illumination control system 10 according to the first embodiment performs setting for automatically narrowing the range for image identification by learning the frequency of the stationary person H until the predetermined time elapses as described above. Yes. On the other hand, in addition to the above setting, the lighting control system 10 of the present embodiment performs different setting processing different from the normal operation on the image processing unit 2 as shown in FIG. A specific range for detecting the presence of the person H is set by causing the specific marker M to be image-identified in the operating state.

  That is, for example, the lighting control system 10 starts up a processing mode for performing another processing of the image processing unit 2 when the lighting fixture 7 is turned on, and the human detection sensor unit 5 performs a predetermined period of time. The operation state is forcibly shifted to a setting operation state that defines a specific range for detecting the presence of the person H. Similarly, the illumination control system 10 may be shifted to an operation state set to determine a specific range for detecting the presence of the person H by an external signal input means such as a remote control device (not shown).

  The lighting control system 10 detects and stores a specific marker M having a predetermined shape arranged in advance in the vicinity of a place where a specific range or the like for detecting the presence of the person H is to be detected during an operation state with a different setting. Let

  That is, the image data from the imaging device 1 is input to the image processing unit 2, but when a specific marker M set in advance is detected during the operation state of another setting, the marker M after the detection is detected. Is set to a specific range for performing the image identification process for detecting the presence of the person H or a specific range for performing the image identification process for detecting the presence of the person H. Then, the lighting control system 10 ends the set operation state by passing a predetermined end time set in advance or by inputting a separate end signal with a remote control device or the like, and shifts to a normal operation state. To do.

  For example, when the lighting control system 10 according to the present embodiment is used for an application such as an office, when the office layout is changed in the predetermined range of the image, a specific range of image identification processing for detecting the presence of the person H is performed. Even when only a part of them is to be canceled, the setting can be changed relatively easily. When attention is paid to applications such as offices, the region where the person H rests in the image viewed from the ceiling surface CS side is limited to the periphery of the desk 21 and the chair. Therefore, the lighting control system 10 arranges a specific marker M having a predetermined shape arranged in the vicinity of a place where a specific range for detecting the presence of the person H is desired, and the area where the person H rests in advance. By acquiring the information, it becomes possible to increase the detection accuracy of the person H.

(Embodiment 3)
As shown in FIG. 7A, the lighting control system 10 of the present embodiment is replaced with a plurality (in this case), instead of the lighting control system 10 that controls one lighting fixture 7 of the first embodiment shown in FIG. The difference is that the lighting control system 10 is constructed by connecting three lighting fixtures 7 to one control device 12. In addition, the same code | symbol is attached | subjected to the component similar to Embodiment 1, and description is abbreviate | omitted suitably.

  The lighting control system 10 of this embodiment is configured by electrically connecting one control device 12 and a plurality of lighting fixtures 7. As shown in FIG. 7B, the control device 12 is capable of outputting a dimming signal for dimming control of the human detection sensor unit 5 and the fluorescent lamp as the light source unit 7a provided in the lighting fixture 7. And a signal output unit 8. Moreover, each lighting fixture 7 is provided with the ballast 7b which functions as the control part 6 which can light-control the light source part 7a. The ballast 7b performs dimming control on the light source unit 7a based on the dimming signal.

  Hereinafter, image processing in the illumination control system 10 of the present embodiment will be described using the captured image of FIG.

  In FIG. 8, the arrangement of a plurality of (here, six) lighting fixtures 7 is illustrated as being superimposed on the captured image generated by imaging with the imaging device 1. Note that four desks 21 are arranged in the captured image.

  In the human detection sensor unit 5, the arithmetic control unit 3 continues the image processing during the lighting control processing E described above. In the second storage unit 4b for preliminarily storing contour shape data such as the person H in the human detection sensor unit 5, the count value stored in the storage area corresponding to each specific range of the lattice shape changes. To go. After a predetermined time has elapsed, the image processing unit 2 compares the count value stored in the second storage unit 4b with a predetermined frequency set in advance as a threshold value. As a result of the comparison, the image processing unit 2 detects a specific range R1 narrower than the entire predetermined range and lower than a predetermined frequency, and a specific range R2 having a predetermined frequency or more. After the predetermined time has elapsed, the image processing unit 2 performs image identification on the result of the feature extraction performed in the specific range R2 having a predetermined frequency or more.

  As shown in FIG. 8, the image processing unit 2 has a specific range R2 (regions a5, a6, b6 and b5 in FIG. 8, regions a9, a10, a11, b11, b10 and b9, f5, f6, The regions of f7, g7, h7, h6, h5, g5, and g6, and the regions of g9, g10, g11, h11, h10, and h9) are defined as regions for performing image processing on the result of feature extraction.

  The image processing unit 2 performs the image difference processing A and the difference pixel extraction processing B as the image processing after the predetermined time has elapsed, and performs the human H (motion) detection processing F over the entire predetermined region in the captured image. . Further, the image processing unit 2 specifies that the number of times that the presence of the person H is detected during the predetermined time exceeds the predetermined frequency for the human H (stationary) detection process D after the recognition process C for the difference pixel. Control is performed only in the range R2.

  The illumination control system 10 of the present embodiment individually controls the illumination light from each illumination fixture 7 based on the detection of the human detection sensor unit 5. More specifically, for example, the lighting control system 10 irradiates the periphery of the specific range R2 with the specific range R2 as a region β (see the dashed line in FIG. 8) where the office work of the person H is performed. The output is set to be relatively high (for example, 70% to 100% with respect to the maximum output of the lighting fixture 7) so that predetermined bright illumination light can be emitted from the lighting fixture 7. Further, in the lighting control system 10, in the arrangement of the lighting fixture 7 in FIG. 8, the lighting fixture 7 that irradiates the region α (see the two-dot chain line in FIG. 8) only in the specific range R1 other than the specific range R2. The output is set to be relatively low (for example, about 25% with respect to the maximum output of the luminaire 7) so that illumination light of a level that can ensure the safety of the passage of the person H can be irradiated. Thereby, the illumination control system 10 of this embodiment is required brightness in the area | region where the office work of the person H is performed compared with what irradiates illumination light with the same output from all the several lighting fixtures 7 equally. It is possible to save energy while ensuring the safety.

  That is, the lighting control system 10 of the present embodiment makes the control of the plurality of lighting fixtures 7 different between the area where the person H is stationary and performs office work and the area where the person H moves for traffic. It is possible to achieve effective energy saving while ensuring the necessary brightness.

H person M marker S1 specific area 1 imaging device 2 image processing unit 5 human detection sensor unit 6 control unit 7a light source unit 10 illumination control system

Claims (6)

Human detection sensor unit including an imaging device and an image processing unit that detects the presence of a person and outputs a detection signal by performing image processing on an image based on image data generated by imaging a specific area by the imaging device And a control unit that controls illumination light emitted from the light source unit based on a detection signal from the human detection sensor unit,
The image processing unit, as the image processing, a first image processing for detecting the presence of a person by performing image identification processing on a result of feature extraction performed over a predetermined range of the image;
The predetermined range is divided into a plurality of different specific ranges, and the number of times the presence of a person is detected for each specific range is counted during a predetermined time. After the predetermined time has elapsed, the predetermined range is counted. And a second image processing for detecting the presence of a person in the specific range that is narrower than the entire range and whose counted frequency exceeds a predetermined frequency.   When the human detection sensor unit detects a specific marker set in advance by performing image identification processing within the predetermined range, the human detection sensor unit detects the presence of a person in the area where the marker is detected after the detection. 2. The illumination according to claim 1, wherein a specific range for performing image identification processing or a region in which the marker is detected is outside a specific range for performing image identification processing for detecting the presence of a person. Control system.   The control unit is connected to a plurality of the light source units, and the control content of the illumination light emitted from the light source unit is a specific range for performing image identification processing for detecting the presence of a person and other than the specific range The illumination control system according to claim 1, wherein the illumination control system is different from each other.   The illumination control system according to claim 3, wherein the control content is brightness of illumination light emitted from each of the plurality of light source units.   The control content is that the control unit starts controlling the illumination light emitted from the plurality of light source units to a predetermined brightness, and then the illumination light emitted from each of the light source units is the predetermined The lighting control system according to claim 3, wherein the time is a time until the brightness is reached.   The control content includes the brightness of the illumination light emitted from each of the plurality of light source units, and the control unit starts controlling the illumination light emitted from the light source unit to have a predetermined brightness. The illumination control system according to claim 3, wherein the illumination light emitted from each of the light source units is a time until the illumination light reaches the predetermined brightness.

Images