JP6058720B2 - Information output device, detection device, program, and information output method - Google Patents

Description

  Embodiments described herein relate generally to an information output device, a detection device, a program, and a method.

  Conventionally, there is a detection device that detects an object (moving body) such as a human body using an image captured by an imaging device such as a monitoring camera. In such a detection apparatus, a difference in luminance between an image representing the current state captured by the imaging apparatus and a reference image (for example, an image several frames before captured by the imaging apparatus) is calculated in units of pixels, and a difference value is calculated. The moving object is detected based on the position of the pixel where becomes equal to or greater than a predetermined value.

  By the way, when a window for light removal is provided in a room such as an office, external light or the like incident from the window may cause a luminance change, and thus may be erroneously detected as a moving object in the above-described detection device. . For this reason, conventionally, there has been proposed a technique for reducing erroneous detection by setting a location where erroneous detection occurs as a non-detected image area (mask).

JP 2011-28956 A JP 2003-324726 A

  However, there is a possibility that the conventional false detection reduction technology cannot cope with various false detection occurrence factors.

The information output apparatus according to the embodiment is an information output apparatus that outputs information related to an image obtained by photographing a subject, and includes a dividing unit, a feature amount calculating unit, a detecting unit, a specifying unit, and an output unit. The dividing unit divides the image into a plurality of image areas. The feature amount calculating means calculates a feature amount that is a difference between the images from a plurality of images having different shooting times for each image area. The detection unit detects an image area having a change in the feature amount based on the feature amount of each image area calculated by the feature amount calculation unit. The specifying unit specifies an image area to be excluded from information output as an excluded area from a plurality of image areas. The output means outputs information relating to the image area obtained by removing the exclusion area specified by the specifying means from the image area detected by the detecting means. Further, the specifying means specifies an image area to be excluded , which is associated with time or time , as an excluded area when excluded by the exclusion setting information .

  According to the present invention, an image area corresponding to a condition determined according to an environment where a subject is placed is specified, and information regarding the image area obtained by removing the specified image area from an image area whose feature amount has changed is output. Thus, it becomes possible to cope with various erroneous detection occurrence factors.

FIG. 1 is a diagram schematically illustrating a configuration of a detection system according to the first embodiment. FIG. 2 is a block diagram showing a configuration of the information output process according to the first embodiment. FIG. 3 is a diagram illustrating an installation example of the information output process according to the first embodiment. FIG. 4 is a block diagram illustrating a functional configuration of the information output process according to the first embodiment. FIG. 5A is a diagram illustrating an example of image division by the image area dividing unit. FIG. 5B is a diagram illustrating an example of image division by the image area dividing unit. FIG. 6A is a schematic diagram illustrating an example of the operation of the feature amount calculation unit. FIG. 6B is a diagram for explaining an example of the operation of the feature amount calculation unit. FIG. 6C is a diagram for explaining an example of the operation of the feature amount calculation unit. FIG. 7-1 is a diagram for explaining an example of the operation of the excluded area specifying unit according to the first embodiment. FIG. 7B is a diagram for explaining an example of the operation of the specifying unit according to the first embodiment. FIG. 8 is a block diagram illustrating the configuration of the detection device according to the first embodiment. FIG. 9 is a block diagram illustrating a functional configuration of the detection device according to the first embodiment. FIG. 10 is a flowchart illustrating the procedure of the information output process of the information output process according to the first embodiment. FIG. 11 is a block diagram illustrating a configuration of information output processing according to the second embodiment. FIG. 12A is a diagram schematically illustrating an example of exclusion setting information. 12B is a diagram schematically illustrating the exclusion target area set in the exclusion setting information in FIG. FIG. 13 is a block diagram illustrating a functional configuration of information output processing according to the second embodiment. FIG. 14 is a flowchart illustrating the procedure of the information output process of the information output process according to the second embodiment. FIG. 15 is a diagram schematically illustrating an example of valid / invalid setting information. FIG. 16 is a block diagram illustrating a functional configuration of information output processing according to the third embodiment. FIG. 17 is a flowchart illustrating the procedure of the information output process of the information output process according to the third embodiment. FIG. 18 is a block diagram illustrating a functional configuration of information output processing according to the fourth embodiment. FIG. 19A is a diagram for explaining an example of the operation of the excluded area specifying unit according to the fourth embodiment. FIG. 19-2 is a diagram for explaining an example of the operation of the excluded area specifying unit according to the fourth embodiment. FIG. 20A is a diagram for explaining another example of the operation of the specifying unit according to the fourth embodiment. FIG. 20-2 is a diagram for explaining another example of the operation of the specifying unit according to the fourth embodiment. FIG. 21 is a flowchart illustrating a procedure of information output processing of the information output apparatus according to the fourth embodiment.

  Hereinafter, embodiments of an information output device, a detection device, a program, and an information output method according to the present invention will be described in detail with reference to the accompanying drawings. In the following embodiments, an example in which the information output device, the object detection device, the program, and the method according to the present invention are applied to object detection in a building such as an office building will be described.

[First Embodiment]
FIG. 1 is a diagram schematically illustrating a configuration of a detection system according to the first embodiment. As shown in the figure, the detection system includes a plurality of information output devices 10 and a detection device 40 connected to the information output devices 10 via a network N such as a LAN (Local Area Network). . The number of information output devices 10 connected to the network N is not particularly limited.

  The information output device 10 includes an imaging unit 16 (see FIG. 2), which will be described later, and outputs information related to an image obtained by capturing a predetermined subject to the detection device 40 as evidence. Hereinafter, the information output apparatus 10 will be described.

  FIG. 2 is a block diagram illustrating a configuration of the information output apparatus 10 according to the first embodiment. As shown in FIG. 1, the information output device 10 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage unit 14, a communication unit 15, and an imaging unit 16. . The CPU 11, the ROM 12, the RAM 13, the storage unit 14, the communication unit 15, and the imaging unit 16 are connected to each other via a bus line 17 such as an address bus or a data bus.

  The CPU 11 comprehensively controls the operation of the information output apparatus 10 by executing various computer-readable programs stored in the ROM 12. In addition, the CPU 11 implements each functional unit (see FIG. 4) described later by executing a predetermined program stored in the ROM 12 and the storage unit 14.

  The ROM 12 stores various programs executed by the CPU 11 and various data. The RAM 13 temporarily stores various programs executed by the CPU 11 and various data, and functions as a working image area. The storage unit 14 is configured by a storage medium such as a hard disk drive device, an SSD (Solid State Drive), or a flash memory, and stores various types of information in addition to programs executed by the CPU 11.

  The communication unit 15 controls data communication with the detection device 40 connected through the network N. The imaging unit 16 is an imaging device having an imaging element such as a color CCD (Charge Coupled Device Image Sensor) or a color CMOS (Complementary Metal Oxide Semiconductor). The imaging unit 16 stores the images sequentially captured at a predetermined frame rate (for example, 30 fps) in the RAM 13.

  FIG. 3 is a diagram illustrating an installation example of the information output apparatus 10. In the figure, a room 20 is a room (closed space) such as an office, and its wall surface is provided with an entrance / exit E1 and lighting windows W1, W2. Further, FIG. 2 shows an example in which the information output device 10 is installed on the ceiling 21 of the room 20. The imaging direction and imaging range of the imaging unit 16 included in the information output device 10 are set in advance so that the entire area of the room 20 can be imaged, and the entire area of the room 20 is imaged as a subject.

  The number of information output devices 10 installed in the room 20 is not limited to one, and a plurality of information output devices 10 may be used. In the case of this form, you may set so that the whole region of the room 20 can be imaged by combining the imaging range of each information output device 10.

  Next, the functional unit of the information output apparatus 10 realized by the CPU 11 of the information output apparatus 10 executing programs stored in the ROM 12 and the storage unit 14 will be described with reference to FIG.

  FIG. 4 is a block diagram illustrating a functional configuration of the information output apparatus 10. As shown in the figure, the CPU 11 of the information output device 10 executes a predetermined program stored in the ROM 12 and the storage unit 14 to thereby execute an image acquisition unit 31, an image area division unit 32, and a feature amount calculation unit. 33, an image area detecting unit 34, an excluded area specifying unit 35, an output unit 36, and a learning unit 37.

  The image acquisition unit 31 sequentially acquires images captured by the imaging unit 16 and stored in the RAM 13 in chronological order. The image area dividing unit 32 divides the image into one or a plurality of image areas for each image acquired by the image acquisition unit 31.

  FIGS. 5A and 5B are diagrams illustrating an example of image division by the image area dividing unit 32. FIG. 5A illustrates an example in which the image G11 acquired by the image acquisition unit 31 is divided into rectangular image areas. FIG. 5B illustrates an example in which the image G12 acquired by the image acquisition unit 31 is divided concentrically from the center of the image G12. Note that the shape and the number of divisions of the image area to be divided are not limited to the examples in FIGS. 5A and 5B and can be arbitrarily set. For example, by setting the number of divisions to 1, the entire image may be set as one image area.

  Returning to FIG. 4, the feature amount calculation unit 33 calculates a feature amount indicating the image feature of the image area for each image area divided by the image area division unit 32. Here, the calculated feature amount is used by the image area detecting unit 34 and the excluded area specifying unit 35. Note that the feature amounts used in the image area detection unit 34 and the excluded area specification unit 35 are not necessarily derived from the same index, but may be derived from different indices.

  For example, the feature amount calculation unit 33 continuously in time series includes the image most recently acquired by the image acquisition unit 31 and the image acquired by the image acquisition unit 31 a predetermined time before this image (for example, one frame before). By comparing the luminance values of the image areas existing at the same position between the two images in units of pixels, a difference value of the luminance values is calculated for each pixel. Then, the feature amount calculation unit 33 calculates the number of pixels in which the absolute value of the calculated difference value is equal to or greater than a predetermined threshold (hereinafter referred to as a first threshold) as the feature amount of the image area. Hereinafter, a method for calculating the feature amount will be described with reference to FIGS. 6-1, 6-2, and 6-3.

  6A, 6B, and 6C are diagrams for explaining an example of the operation of the feature amount calculation unit 33. FIG. If the image acquired by the image acquisition unit 31 at time t1 is G21 in FIG. 6A and the image acquired by the image acquisition unit 31 at the latest time t1 + 1 is G22 in FIG. The amount calculation unit 33 compares the luminance values of the image areas existing at the same position in units of pixels between the two images G21 and G22, and calculates the difference.

  Here, if the person H1 in the images G21 and G22 has moved from time t1 to time t1 + 1, the person H1 is displayed due to the difference in the position of the person H1 in the images as shown in FIG. 6-3. The luminance value of the pixel to be changed changes from time t1 to time t1 + 1. Therefore, the feature amount calculation unit 33 calculates the amount of change as the difference between the luminance values, and determines the number of pixels in which the absolute value of the difference value is equal to or greater than the first threshold value for each image area (FIG. Feature value).

  The feature amount calculation method is not limited to the above example, and other calculation methods may be used. For example, for each of the same image areas of two images that are continuous in time series, the sum of the luminance values of each pixel constituting the image area is calculated, and the difference value of the sum is obtained as the feature amount of the image area. May be. Further, for each image area of the image acquired most recently by the feature amount calculation unit 33, the sum of the luminance values of the pixels constituting the image area is calculated, and this sum value is acquired as the feature amount of the image area. Also good.

  In the above example, the difference between the luminance values is used as the feature amount. However, other elements such as a luminance value histogram, luminance distribution, and color information may be used as the feature amount. For example, for each image area of the image most recently acquired by the feature amount calculation unit 33, the human body is identified from the image in the image area by pattern matching or the like, and the humanity value obtained as a result is obtained as a feature of the image area. It may be an amount. Note that the value of the first threshold value is changed according to the feature amount calculation method and the index used as the feature amount.

  Returning to FIG. 4, based on the calculated feature amount of each image area, the image area detection unit 34 may have an object (moving object) such as a person in the image area where the feature amount has changed. Detect as an area. Specifically, the image area detection unit 34 compares the feature amount of each image area calculated by the feature amount calculation unit 33 with a predetermined threshold (hereinafter referred to as a second threshold), and the feature amount is the second threshold. The above image area is detected as an image area in which an object (moving body) such as a human body may exist. For example, in the state of FIG. 6-3 described above, if the feature amount of each image area surrounded by a black frame is equal to or greater than the second threshold value, the image area detection unit 34 determines each image area surrounded by the black frame. Is detected respectively.

  The excluded area specifying unit 35 specifies an image area corresponding to a condition determined according to the environment where the subject is placed, from each image area constituting the entire image. Specifically, the excluded area specifying unit 35 compares the feature amount of each image area constituting the entire image with a predetermined threshold (hereinafter referred to as a third threshold), so that the feature amount is equal to or greater than the third threshold. Specify the image area.

  In the present embodiment, the feature amount (hereinafter referred to as the first feature amount) used by the image area detection unit 34 described above to detect the image area and the feature amount (excluding area specifying unit 35) used for comparison with the third threshold ( Hereinafter, the second feature quantity) can be handled as an independent physical quantity. Therefore, the first feature value and the second feature value may be values obtained by different indices (for example, the difference between the brightness values and the sum of the brightness values), or may be values obtained by the same index. Good. In the present embodiment, the processing target of the exclusion area specifying unit 35 is the entire image. However, the present invention is not limited to this, and the image area detected by the image area detection unit 34 may be the processing target.

  The output unit 36 outputs information regarding the remaining image areas, excluding the image areas specified by the excluded area specifying unit 35, from the image areas detected by the image area detecting unit 34 to the detection device 40 as evidence.

  Here, information output as evidence is not particularly limited. For example, coordinates representing a remaining image area obtained by excluding the image area specified by the excluded area specifying unit 35 from the image area detected by the image area detecting unit 34 may be output as evidence. Further, an image (data) representing the remaining image area excluding the image area specified by the excluded area specifying unit 35 from the image area detected by the image area detecting unit 34 may be output as evidence. Further, an image obtained by superimposing information indicating the image area of the exclusion result on the image captured by the imaging unit 16 may be output as evidence. Further, even if the remaining image area becomes null as a result of the exclusion, information indicating that may be output as evidence.

  Hereinafter, operation | movement of the exclusion area specific | specification part 35 is demonstrated using FIGS. 7-1 and 7-2. FIGS. 7A and 7B are diagrams for explaining an example of the operation of the excluded area specifying unit 35. Here, it is assumed that the image G31 in FIG. 7-1 is an image acquired by the image acquisition unit 31 at time t2, and the image G32 in FIG. 7-2 is an image acquired by the image acquisition unit 31 at the latest time t2 + 1. Suppose there is. Images G31 and G32 show a state in which the indoor lighting is switched from OFF to ON from time t2 to time t2 + 1.

  In the example of FIGS. 7-1 and 7-2, the feature amount calculation unit 33 calculates the sum of the luminance values of the pixels constituting each image area in each of the two images G31 and G32 having different imaging times. Assume that the difference value of the sum in the corresponding image area is calculated as the second feature amount of the image area. For example, assuming that the sum of the brightness values in each image area of the image G31 is “5000” and the sum of the brightness values in each image area of the image G32 is “30000”, the feature amount calculation unit 33 calculates the sum of these sums. The difference value “25000” is calculated as the second feature amount of each image area.

  As described above, when the interior of the room is illuminated by lighting, the luminance value of the illuminated image area changes. Therefore, for example, when the image area detection unit 34 detects an image area based on a change in luminance value (first feature amount), no object exists due to a change in luminance value accompanying lighting. There is a possibility that all or a part of the image area in the image G32 is erroneously detected.

  Therefore, for example, by setting a value corresponding to the difference in luminance value when lighting is turned on / off in the room that is the subject as the third threshold (for example, “20000”) in advance, the excluded area specifying unit 35 The image area in which the second feature amount is equal to or greater than the third threshold is specified from the image. As a result, the output unit 36 excludes the image area specified by the exclusion area specifying unit 35 from the image area detected by the image area detecting unit 34, so that the second of the image areas detected by the image area detecting unit 34. An image area having a feature amount of less than 20000 remains. Therefore, by appropriately setting the value of the third threshold value, it is possible to reduce the occurrence of erroneous detection caused by a change in the luminance value associated with lighting on / off.

  By the way, the third threshold value that is a parameter in the excluded area specifying unit 35 varies depending on the configuration of the room that is the subject, the index as the second feature value, the detection timing (season, date, time), and the like. Therefore, the learning unit 37 performs learning using the processing results of the feature amount calculation unit 33 and the excluded area specifying unit 35 as inputs and the third threshold as an output.

  For example, the learning unit 37 has a plurality of third threshold values and processing results of the excluded area specifying unit 35 when the third threshold values are assigned for each type of element as the feature amount and each detection timing (or The processing result of the output unit 36) is stored in the storage unit 14 in association with each other. Then, the learning unit 37 performs machine learning on a threshold set from which a more appropriate processing result (detection result) is obtained based on the plurality of third threshold values and processing results stored in the storage unit 14, and sets the third threshold value. This is reflected in the third threshold value of the excluded area specifying unit 35.

  In addition, the technique and method which concern on the machine learning in the learning part 37 shall use a well-known technique and method. Moreover, although the parameter used as learning object was made into the 3rd threshold value, you may include not only in this but a 2nd threshold value or a 1st threshold value in the learning object.

  Returning to FIG. 1, based on the evidence input from the information output device 10, the detection device 40 determines whether an object (moving body) such as a person exists in the room that is the subject of each information output device 10. It is a server device.

  FIG. 8 is a block diagram illustrating a configuration of the detection device 40 according to the first embodiment. As shown in the figure, the detection device 40 includes a CPU 41, a ROM 42, a RAM 43, a storage unit 44, an operation unit 45, a display unit 46, and a communication unit 47. The CPU 41, the ROM 42, the RAM 43, the storage unit 44, the operation unit 45, the display unit 46, and the communication unit 47 are connected to each other via a bus line 48 such as an address bus or a data bus.

  The CPU 41 comprehensively controls the operation of each unit of the detection device 40 by developing and executing a predetermined program stored in the ROM 42 or the storage unit 44 in the RAM 43. In addition, the CPU 41 implements a function unit described later by expanding and executing a predetermined program stored in the ROM 42 or the storage unit 44 in the RAM 43.

  The ROM 42 stores various programs executed by the CPU 41 and setting information. The RAM 43 is a main storage device and is used as a work memory.

  The storage unit 44 is an auxiliary storage device such as an HDD (Hard Disk Drive) and stores various programs executed by the CPU 41 and various types of information. The operation unit 45 is an input device such as a keyboard and a mouse, and outputs operation contents received from the operator to the CPU 41. The display unit 46 is a display device such as an LCD (Liquid Crystal Display), and displays characters, images, and the like under the control of the CPU 41.

  The communication unit 47 is a network interface connectable to the network N, and exchanges information with each information output device 10 connected to the network N under the control of the CPU 41.

  FIG. 9 is a block diagram illustrating a functional configuration of the detection device 40. As shown in the figure, the CPU 41 of the detection device 40 executes a predetermined program stored in the ROM 42 and the storage unit 44, thereby providing an evidence reception unit 51, a state determination unit 52, and a state presentation unit 53. It has a function.

  The evidence receiving unit 51 receives the evidence transmitted (output) from each of the information output devices 10 via the communication unit 47. Note that the evidence includes an identifier (for example, an IP address) that can identify each of the information output apparatuses 10.

  Based on the information included in the evidence, the state determination unit 52 determines whether or not an object (moving body) such as a person exists within the jurisdiction (imaging range) of the information output apparatus 10 that has output the evidence. . Specifically, when information such as coordinates representing a specific image area or the image itself is included in the evidence, the state determination unit 52 corresponds to the image area within the jurisdiction of the information output apparatus 10 that has output the evidence. It is determined that an object (moving object) exists at the position. In addition, when the evidence that the image area is null is included in the evidence, the state determination unit 52 determines that there is no object (moving object) within the jurisdiction of the information output apparatus 10 that has output the evidence.

  The state presentation unit 53 presents the display result to the operator of the detection device 40 by outputting the determination result in the state determination unit 52 to the display unit 46 and the like. The evidence received from the information output device 10 and the determination result in the state determination unit 52 based on the evidence may be associated with each other and accumulated and managed in the storage unit 44.

  Next, the operation of the information output apparatus 10 of this embodiment will be described with reference to FIG. Here, FIG. 10 is a flowchart showing a procedure of information output processing of the information output apparatus 10.

  First, the image acquisition unit 31 sequentially acquires images captured by the imaging unit 16 and stored in the RAM 13 in chronological order (step S11). Each time an image is acquired in step S11, the image area dividing unit 32 divides the image into one or a plurality of image areas (step S12).

  The feature amount calculation unit 33 calculates the first feature amount and the second feature amount of the image area for each image area divided in step S12 for the image most recently acquired in step S11 (step S13). Subsequently, the image area detection unit 34 compares the first feature value of each image area calculated in step S13 with the second threshold value, and detects an image area whose first feature value is equal to or greater than the second threshold value ( Step S14). Next, the excluded area specifying unit 35 compares the second feature amount of each image area calculated in step S13 with the third threshold value, and specifies an image area whose second feature amount is equal to or greater than the third threshold value (step S13). S15).

  Subsequently, the output unit 36 excludes the image area detected in step S15 from the image area detected in step S14 (step S16). And the output part 36 outputs the information regarding the remaining image area which becomes the processing result in step S16 as evidence to the detection device 40 (step S17), and returns to step S11 again.

  As a result, in the detection device 40, the evidence receiving unit 51 receives the evidence output by the information output device 10, and based on the evidence, the state determination unit 52 outputs the evidence within the jurisdiction of the information output device 10. It is determined whether or not an object exists. And the state presentation part 53 presents the determination result in the state determination part 52. FIG.

  As described above, according to the first embodiment, when the second feature amount of the image area constituting the image deviates from the third threshold value determined according to the environment of the room (space) that is the subject, The image area is specified and excluded from the detection result of the image area detection unit 34. As a result, for example, it is possible to reduce erroneous detection caused by a change in luminance value associated with turning on / off of illumination, thereby improving detection accuracy.

  By the way, the configuration of the information output apparatus 10 can adopt various forms. For example, when pixel color information is used as the index of the second feature value, the third threshold value is the vicinity of whiteout (such as 90% or more of the maximum value of each RGB component) or the color of sunset (R component is the maximum value). 90% or more) can be set to reduce the occurrence of false detection caused by light incident from a window or the like. In the above example, the image area having the second feature amount equal to or greater than the third threshold is removed from the detection result. However, the image area in which the second feature amount is within (or out of) the preset threshold range. May be removed from the detection result.

  Moreover, in the said embodiment, although the detection apparatus 40 was set as the structure provided with the function of the state determination part 52, not only this but the information output device 10 is good also as a structure provided with the function of the state determination part 52. In the case of this configuration, the determination result in the state determination unit 52 may be included in the evidence and output.

[Second Embodiment]
Next, an information output apparatus according to the second embodiment will be described. In addition, about the component similar to 1st Embodiment, the same code | symbol is provided and description is abbreviate | omitted.

  FIG. 11 is a block diagram illustrating a configuration of an information output apparatus 10a according to the second embodiment. As shown in the figure, the information output device 10a includes a CPU 11, a ROM 12, a RAM 13, a storage unit 14a, a communication unit 15, and an imaging unit 16. The CPU 11, the ROM 12, the RAM 13, the storage unit 14a, the communication unit 15, and the imaging unit 16 are connected to each other via a bus line 17 such as an address bus or a data bus.

  The storage unit 14a corresponds to the storage unit 14 described above, and stores various types of information in addition to programs executed by the CPU 11. Further, the storage unit 14a stores, as one of various types of information, exclusion setting information 141 that associates an image area to be excluded from the detection target and a time (date and time) at which the image area is excluded from the detection target. .

  Here, FIG. 12A is a diagram schematically illustrating an example of the exclusion setting information 141. FIG. 12B is a diagram schematically illustrating the exclusion target area set in the exclusion setting information 141 in FIG.

  As illustrated in FIG. 12A, the exclusion setting information 141 includes data on an exclusion target area 141a in which an image area to be excluded from a detection target is stored and an exclusion time 141b in which a time to be excluded from the detection target is stored. It has as an item. Hereinafter, the image area stored (set) in the exclusion target area 141a is referred to as “exclusion target area”.

  By storing the coordinates of a plurality of image areas together in the exclusion target area 141a, an image area group composed of a plurality of image areas can be handled as one exclusion target area. For example, the exclusion target areas A11, A12, and A13 stored in the exclusion target area 141a in FIG. 12A correspond to the image areas A11, A12, and A13 (black frame portions) illustrated in FIG. In addition, it is good also as a form which overlaps and sets one part or all the image areas by some exclusion object areas.

  The exclusion time 141b stores a time at which exclusion from the detection target starts and a time at which the exclusion ends. Note that the start / end timing of exclusion may be specified using not only the time but also the date (month, day), day of the week, or the like. Moreover, it is good also as a form which always designates exclusion from a detection target by making the exclusion time 141b blank (null) like the exclusion time 141b linked | related with exclusion object area A12. A plurality of different exclusion times may be set for the same exclusion target area.

  By the way, it is preferable to set values according to the indoor environment that is the subject in the exclusion target area 141a and the exclusion time 141b described above. For example, in the image area A11 shown in FIG. 12-2, in an environment in which the morning sun enters between 6:00 AM and 8:00 AM, the image area A11 is set as an exclusion target area, and For example, “06: 0 to 08:00” is set as the exclusion time. As a result, even when the luminance value of the image area A11 changes with the incidence of the morning sun, the image area A11 can be excluded from the detection target, so that an image area in which no object such as a human body exists is erroneously detected. Can be reduced.

  FIG. 13 is a block diagram illustrating a functional configuration of the information output apparatus 10a according to the second embodiment. As shown in the figure, the CPU 11 of the information output device 10a executes a predetermined program stored in the ROM 12 and the storage unit 14a, whereby an image acquisition unit 31, an image area division unit 32, and a feature amount calculation unit. 33, an image area detecting unit 34, an excluded area specifying unit 35a, an output unit 36, a learning unit 37a, and a time measuring unit 38.

  The timer 38 generates current time information representing the current date and time based on the time measured by a timer such as RTC (Real Time Clock) provided in the information output device 10a.

  The exclusion area specifying unit 35a refers to the exclusion setting information 141, and specifies an exclusion target area corresponding to the current date and time of the current time information generated by the timing unit 38.

  The learning unit 37a has the function of the learning unit 37 described above, and performs learning by using the processing result of the exclusion area specifying unit 35a as an input and outputting the exclusion target area and the exclusion time.

  For example, the learning unit 37a may include an exclusion target area set around an arrangement position of a false detection factor (such as a window) existing in the room, various exclusion times assigned to the exclusion target area, and exclusion under this condition. The processing result of the area specifying unit 35a (or the processing result of the output unit 36) is associated and stored in the storage unit 14a. Then, the learning unit 37a performs machine learning of an exclusion time at which a more appropriate processing result (detection result) is obtained based on the combination of the exclusion target area, the exclusion time, and the processing result stored in the storage unit 14a. The exclusion time is reflected (stored) in the exclusion time 141b of the exclusion setting information 141.

  It should be noted that known techniques and methods are used as techniques and methods related to machine learning in the learning unit 37a. Further, an exclusion target area may be included as a learning target.

  Next, the operation of the information output apparatus 10a of this embodiment will be described with reference to FIG. Here, FIG. 14 is a flowchart showing a procedure of information output processing of the information output apparatus 10a. Note that steps S21 to S24 are the same as steps S11 to S14 of the first embodiment described above, and a description thereof will be omitted.

  In step S25, the exclusion area specifying unit 35a refers to the exclusion setting information 141, and specifies an exclusion target area corresponding to the current date and time of the current time information generated by the timing unit 38 (step S25).

  Subsequently, the output unit 36 excludes the exclusion target area specified in step S25 from the image area detected in step S24 (step S26). And the output part 36 outputs the information regarding the remaining image area which becomes the processing result in step S26 as evidence to the detection device 40 (step S27), and returns to step S21 again.

  As a result, in the detection device 40, the evidence receiving unit 51 receives the evidence output from the information output device 10 a, and based on this evidence, the state determination unit 52 outputs the evidence within the jurisdiction of the information output device 10. It is determined whether or not an object exists. And the state presentation part 53 presents the determination result in the state determination part 52. FIG.

  As described above, according to the second embodiment, an exclusion target area in which the time when an image is captured and the position of each image area constituting the image are determined according to the environment of the room (space) that is the subject. And when it corresponds to the conditions of exclusion time, the exclusion object area corresponding to this condition is specified, and it excludes from the detection result in the image area detection part 34. FIG. Thereby, for example, it is possible to reduce erroneous detection caused by light incident from the window, shaking of a curtain attached to the window, and the like, and thus detection accuracy can be improved. In addition, since the exclusion target area can be validated / invalidated based on the indoor usage environment (for example, business days and closed days) as a subject, it is possible to flexibly cope with the indoor usage environment.

  By the way, the configuration related to the exclusion setting information 141 in the information output apparatus 10a can adopt various forms. For example, in the second embodiment, by specifying the exclusion target area of the exclusion time corresponding to the current date and time, the exclusion target area is excluded from the detection result in the image area detection unit 34. For each target area, setting information (hereinafter referred to as valid / invalid setting information) that specifies validity / invalidity is separately provided, and exclusion / exclusion of exclusion target areas is valid / invalid based on the valid / invalid setting information It is good also as a structure to make.

  Here, FIG. 15 is a diagram schematically illustrating an example of the valid / invalid setting information 142. As shown in FIG. 15, the valid / invalid setting information 142 includes a processing target area 142a in which an exclusion target area to be processed is stored, and a valid / invalid setting 142b in which data defining valid / invalid is stored. As data items. The valid / invalid setting information 142 is stored in the storage unit 14a.

  By storing the same exclusion target area as the exclusion setting information 141 in the processing target area 142a, the exclusion setting information 141 and the valid / invalid setting information 142 are associated with each other using the exclusion target area as a key. The exclusion target area 141a and the processing target area 142a may be made common so that the exclusion setting information 141 and the valid / invalid setting information 142 are merged to form one setting information.

  When using the valid / invalid setting information 142 described above, the exclusion area specifying unit 35a determines whether the validity / invalidity setting information 142 associated with the exclusion target area corresponding to the current date and time is valid prior to the execution of step S25. Determine. Then, when the valid / invalid setting information 142 is “valid”, the excluded area specifying unit 35a proceeds to the above-described step S25 and specifies the excluded area corresponding to the current date and time. When the valid / invalid setting information 142 is “invalid”, the excluded area specifying unit 35a proceeds to step S26 without specifying the excluded area corresponding to the current date and time.

  Thereby, since the function of the exclusion area specifying unit 35a can be validated / invalidated in units of exclusion target areas stored in the exclusion setting information 141, exclusion of the exclusion target areas can be individually performed in units of exclusion target areas. Can be controlled.

  The valid / invalid setting for the exclusion target area is set in the exclusion setting information 141, but is not limited to this. When the image area detection unit 34 detects the image area, the detection is detected. A valid / invalid setting for the image area may be received from an external device (for example, the detection device 40), and the exclusion area identifying unit 35a may identify the exclusion target area based on the setting content.

[Third Embodiment]
Next, an information output apparatus according to the third embodiment will be described. In addition, about the component similar to 1st Embodiment, the same code | symbol is provided and description is abbreviate | omitted.

  FIG. 16 is a block diagram illustrating a functional configuration of the information output apparatus 10b according to the third embodiment. As shown in the figure, the CPU 11 of the information output device 10b executes a predetermined program stored in the ROM 12 and the storage unit 14, thereby executing an image acquisition unit 31, an image area division unit 32, and a feature amount calculation unit. 33, an image area detecting unit 34, an excluded area specifying unit 35b, an output unit 36, and a learning unit 37b. It is assumed that the hardware configuration of the information output device 10b is the same as that of the first embodiment.

  For each image area of a plurality of images (at least three or more) having different shooting times, the excluded area specifying unit 35b compares the second feature amount of the image area in chronological order, and the second feature amount changes regularly. The image area to be specified is specified.

  By the way, the state in which the second feature amount regularly changes is, for example, when a periodic image (video) such as a screen saver is displayed on a PC monitor provided in the room, or when the fan of a fan is rotating. This occurs when an object that performs regular motion is included in the image. In such a case, if the detection result in the image area detection unit 34 is output as it is, an image area in which an object such as a human body does not exist is output as the detection result. Therefore, in the present embodiment, the excluded area specifying unit 35b specifies an image area in which the second feature amount regularly changes, and the output unit 36 detects the specified image area by the image area detecting unit 34. By removing from the result, it is possible to reduce an event that a regular operation is erroneously detected as an object.

  As described above, the excluded area specifying unit 35b specifies an image area to be excluded based on the regularity (pattern) of the change in the second feature amount of each image area between a plurality of images having different shooting times. . It is assumed that the parameters related to regular change determination can be arbitrarily set. Parameters relating to regular change determination include, for example, the index of the second feature, the periodic determination period, the number of images used for determination, the sampling interval, and the position of the image area to be determined (Entrance / exit E1, the vicinity of the window W1, etc.), the time (date and time) for executing the determination, and the like.

  The learning unit 37b has the function of the learning unit 37 described above, and performs learning by using the processing result of the excluded area specifying unit 35b as an input and outputting a parameter related to regular change determination.

  For example, the learning unit 37b associates various parameter sets related to regular change determination with the processing results of the exclusion area specifying unit 35b (or processing results of the output unit 36) in each parameter set, and stores the storage unit 14 Remember it. Then, the learning unit 37b performs machine learning on a parameter set from which a more appropriate processing result (detection result) is obtained based on the set of the parameter set stored in the storage unit 14 and the processing result, and excludes the parameter set. This is reflected in the area specifying unit 35b. In addition, the technique and method which concern on the machine learning in the learning part 37b shall use a well-known technique and method.

  Next, the operation of the information output apparatus 10b of this embodiment will be described with reference to FIG. Here, FIG. 17 is a flowchart showing a procedure of information output processing of the information output apparatus 10b. Note that steps S31 to S34 are the same as steps S11 to S14 of the first embodiment described above, and a description thereof will be omitted.

  In step S35, the excluded area specifying unit 35b compares the second feature amount of the image area in time series for each image area of a plurality of images having different shooting times, and the second feature amount changes regularly. An image area is specified (step S35).

  Subsequently, the output unit 36 excludes the image area specified in step S35 from the image area detected most recently in step S34 (step S36). And the output part 36 outputs the information regarding the remaining image area which becomes the processing result in step S36 as evidence to the detection device 40 (step S37), and returns to step S31 again.

  As a result, in the detection device 40, the evidence receiving unit 51 receives the evidence output by the information output device 10b, and based on this evidence, the state determination unit 52 outputs the evidence within the jurisdiction of the information output device 10 It is determined whether or not an object exists. And the state presentation part 53 presents the determination result in the state determination part 52. FIG.

  As described above, according to the third embodiment, the second feature amount of each image area constituting the image corresponds to the regular change condition determined according to the environment of the room (space) that is the subject. In this case, an image area corresponding to this condition is specified and excluded from the detection result of the image area detection unit 34. As a result, for example, it is possible to reduce false detection caused by an object that performs a regular operation such as a periodic image displayed on a PC monitor provided in a room or a fan of a fan. Can be improved.

[Fourth Embodiment]
Next, an information output apparatus according to the fourth embodiment will be described. In addition, about the component similar to 1st Embodiment, the same code | symbol is provided and description is abbreviate | omitted.

  FIG. 18 is a block diagram illustrating a functional configuration of the information output apparatus 10c according to the fourth embodiment. As shown in the figure, the CPU 11 of the information output device 10c executes a predetermined program stored in the ROM 12 and the storage unit 14, whereby an image acquisition unit 31, an image area division unit 32, and a feature amount calculation unit. 33, an image area detecting unit 34, an excluded area specifying unit 35c, an output unit 36, and a learning unit 37c. Note that the hardware configuration of the information output apparatus 10c is the same as that of the first embodiment.

  The exclusion area specifying unit 35c compares the positions of the image areas detected by the image area detection unit 34 in time series for each of a plurality of images having different imaging times, so that the locus of the position continuously moves. It is determined whether or not the image area exists in a discontinuous position other than the image area that is continuously moving.

  19A and 19B are diagrams for explaining an example of the operation of the excluded area specifying unit 35c. Here, it is assumed that the image G41 of FIG. 19-1 is an image acquired by the image acquisition unit 31 at time t3, and the image G42 of FIG. 19-2 is an image acquired by the image acquisition unit 31 at time t3 + 1. To do. In addition, it is assumed that an entrance E1 to the building exists at a position P11 in the figure, and windows W1 and W2 exist at positions P12 and P13 in the figure (see FIG. 3).

  Assume that the image area detection unit 34 detects the image area A42 (black frame portion) from each image area of the image G42 at time t3 + 1 under the above conditions. In this case, the excluded area specifying unit 35c specifies the position of the image area A42 from the image G42 by setting the position of the image area in the image G41 to be null. Then, the excluded area specifying unit 35c determines that the image area A42 is suddenly detected at time t3 + 1 by comparing the positions of the image areas specified from the images G41 and G42 in time series.

  In the case of the examples of the images G41 and G42, for example, a situation in which the light incident from the window W1 of P12 in the figure fluctuates at time t3 + 1 can be considered. In this case, the image area A42 detected by the image area detection unit 34 is erroneously detected because there is no object such as a person. Therefore, the excluded area specifying unit 35c specifies the image area A42 that exists in a discontinuous position, that is, suddenly appears in the two images G41 and G42 having different imaging times.

  FIGS. 20A and 20B are diagrams for explaining another operation example of the excluded area specifying unit 35c. Here, it is assumed that the image G51 of FIG. 20-1 is an image acquired by the image acquisition unit 31 at time t4, and the image G52 of FIG. 20-2 is an image acquired by the image acquisition unit 31 at time t4 + 1. To do. As in FIG. 19A, it is assumed that the entrance E1 into the building exists at the position P11 in the drawing, and the windows W1 and W2 exist at the positions P12 and P13 in the drawing (see FIG. 3).

  Under the above conditions, at time t4, the image area detection unit 34 detects the image area A51 (black frame portion) from each image area shown in the image G51, and at time t4 + 1, the image from each image area shown in the image G52. Assume that an area A52 (black frame portion) is detected. In this case, the excluded area specifying unit 35c specifies the position of the image area A51 from the image G51 and specifies the position of the image area A52 from the image G52. And the exclusion area specific | specification part 35c compares the position of image area A51, A52 in time series order, and since the image area A52 exists in the vicinity of the position of image area A51, when the said position is moving continuously, judge.

  In the case of the examples of the images G51 and G52, it can be considered that a person who has entered the room through the entrance E1 has moved from the image area A51 to the image area A52 from time t4 to time t4 + 1. In this case, the image area A52 detected by the image area detector 34 is normally detected because an object such as a person exists. For this reason, the excluded area specifying unit 35c does not specify the image areas A51 and A52, and leaves them as the detection result of the image area detecting unit 34.

  As described above, the excluded area specifying unit 35c specifies an image area to be excluded based on the continuity of the position of the image area detected by the image area detecting unit 34 between a plurality of images having different shooting times. Note that it is possible to arbitrarily set parameters relating to the determination of continuity. Parameters relating to the determination of continuity include, for example, an index used as the second feature amount, the number of images used for determination of continuity, the distance between image areas determined to be continuous, the sampling interval, and the like. Examples include the image area position (the entrance / exit E1, the periphery of the window W1, etc.), the time (date and time) at which the determination is performed, and the like.

  Among the above examples, the distance between image areas determined to be continuous may be defined as, for example, image areas located in the vicinity of 4 or 8 in the lattice-shaped image area being continuous, You may define that the image area expanded by the image area is continuous. Further, the number of images used for determination of continuity is not limited to two, and determination may be made using three or more images.

  In addition, when the image area position to be subjected to the continuity determination is defined, the existence position of the entrance / exit (P11) and the windows (P12, P13) of the building is stored in advance as setting information, and the image area detection unit 34 The continuity is determined in consideration of whether the image area detected by is present in the vicinity of the entrance E1 or the windows W1 and W2. For example, when an image area is detected near the entrance E1, the image area may be excluded from the continuity determination target.

  The learning unit 37c has the function of the learning unit 37 described above, and performs learning using the processing result in the excluded area specifying unit 35c as an input and a parameter related to determination of continuity as an output.

  For example, the learning unit 37c associates various parameter sets related to determination of continuity with the processing results of the excluded area specifying unit 35c (or processing results of the output unit 36) in each parameter set in the storage unit 14. Remember. Then, the learning unit 37c performs machine learning on the parameter set that can obtain a more appropriate processing result (detection result) based on the combination of the parameter set stored in the storage unit 14 and the processing result, and excludes the parameter set. This is reflected in the area specifying unit 35c. It should be noted that known techniques and methods are used as techniques and methods related to machine learning in the learning unit 37c.

  Next, the operation of the information output apparatus 10c of this embodiment will be described with reference to FIG. Here, FIG. 21 is a flowchart showing a procedure of information output processing of the information output apparatus 10c. Note that steps S41 to S44 are the same as steps S11 to S14 of the first embodiment described above, and a description thereof will be omitted.

  In step S45, the excluded area specifying unit 35c compares the positions of the image areas detected in step S44 in time series for each image area of a plurality of images having different shooting times, and images present at discontinuous positions. An area is specified (step S45).

  Subsequently, the output unit 36 excludes the image area specified in step S45 from the image area detected most recently in step S44 (step S46). And the output part 36 outputs the information regarding the remaining image areas which become the processing result in step S46 as evidence to the detection device 40 (step S47), and returns to step S41 again.

  As a result, in the detection device 40, the evidence receiving unit 51 receives the evidence output by the information output device 10 c, and based on this evidence, the state determination unit 52 outputs the evidence within the jurisdiction of the information output device 10. It is determined whether or not an object exists. And the state presentation part 53 presents the determination result in the state determination part 52. FIG.

  As described above, according to the fourth embodiment, in each image area constituting the image, the position of the image area detected by the image area detection unit 34 depends on the environment of the room (space) that is the subject. When the discontinuous position is not met, the image area is specified and excluded from the detection result of the image area detection unit 34. Thereby, for example, it is possible to reduce erroneous detection caused by light incident from the window, shaking of a curtain attached to the window, and the like, and thus detection accuracy can be improved.

  As mentioned above, although embodiment of this invention was described, the said embodiment was shown as an example and is not intending limiting the range of invention. The above embodiment can be implemented in various other forms, and various omissions, replacements, changes, combinations, additions, and the like can be made without departing from the scope of the invention. Moreover, the said embodiment and its deformation | transformation are included in the range of the invention, the summary, and the invention described in the claim, and its equal range.

  For example, in the embodiment described above, the information output device 10 (10a, 10b, 10c) has been described as being attached to the ceiling 21.

  Moreover, in the said embodiment, although the information output device 10 (10a, 10b, 10c) was set as the form which incorporated the imaging part 16, it is good also as a form which attaches the imaging part 16 not only to this. In this case, only the imaging unit 16 may be attached to the ceiling 21 and other elements may be realized by an information processing device such as the detection device 40, for example.

  In the above embodiment, the detection target space is a closed space such as the room 20. However, the present invention is not limited thereto, and may be an open space such as a park or a road.

  In the above embodiment, the information output device 10 (10a, 10b, 10c) is configured to learn parameters related to image area specification. However, the present invention is not limited to this, and an external device such as the detection device 40 learns. The function of the unit 37 (37a, 37b, 37c) may be provided, and the parameters obtained by learning may be reflected in the information output device 10 (10a, 10b, 10c).

  In addition, the program executed by the information output device 10 (10a, 10b, 10c) of the above embodiment is provided by being incorporated in advance in a storage medium (ROM 12 or storage unit 14) included in each device. The file is not limited to an installable or executable format, and is provided by being recorded on a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, or a DVD (Digital Versatile Disk). You may comprise as follows. Furthermore, the storage medium is not limited to a medium independent of a computer or an embedded system, but also includes a storage medium in which a program transmitted via a LAN, the Internet, or the like is downloaded and stored or temporarily stored.

  Further, the program executed by the information output apparatus 10 (10a, 10b, 10c) of the above embodiment is stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. Alternatively, it may be configured to be provided or distributed via a network such as the Internet.

10, 10a, 10b, 10c Object detection device 11 CPU
12 ROM
13 RAM
14, 14a Storage unit 141 Exclusion setting information 142 Invalid setting information 15 Communication unit 16 Imaging unit 31 Image acquisition unit 32 Image area division unit 33 Feature amount calculation unit 34 Image area detection unit 35, 35a, 35b, 35c Exclusion area specification unit 36 Output unit 37, 37a, 37b, 37c Learning unit 38 Timekeeping unit 40 Detection device 41 CPU
42 ROM
43 RAM
44 storage unit 45 operation unit 46 display unit 47 communication unit 51 evidence reception unit 52 state determination unit 53 state presentation unit

Claims (11)

An information output device that outputs information about an image of a subject,
Dividing means for dividing the image into a plurality of image areas;
Feature amount calculating means for calculating, for each image area, a feature amount that is a difference between the images from a plurality of images having different shooting times;
Detecting means for detecting an image area having a change in the feature amount based on the feature amount of each image area calculated by the feature amount calculating unit;
A specifying means for specifying an image area to be excluded from information output as an excluded area from the plurality of image areas;
Output means for outputting information on an image area obtained by removing the exclusion area specified by the specifying means from the image area detected by the detecting means;
With
The information output device that specifies the image area to be excluded, which is associated with time or time , as the exclusion area when the identification means is excluded by the exclusion setting information .   The information output apparatus according to claim 1, wherein the specifying unit specifies an image area in which a feature amount of the image area deviates from a predetermined threshold as the exclusion area.   The information output apparatus according to claim 1, wherein the specifying unit specifies an image area in which a locus of the position of the image area detected by the detecting unit is discontinuous as the exclusion area.   The information output device according to any one of claims 1 to 3, wherein the specifying unit uses one or a plurality of image areas as a processing unit and specifies the exclusion area for each processing unit. Storage means for storing setting information defining validity / invalidity for each processing unit;
The information output device according to claim 4, wherein the specifying unit specifies the exclusion area for a processing unit defined as valid in the setting information.   The information output according to any one of claims 1 to 3, further comprising learning means for performing learning using, as an input, a processing result of the specifying means and outputting a parameter relating to the specification of the exclusion area by the specifying means. apparatus. A dividing unit that divides an image of a subject into a plurality of image areas;
Feature amount calculating means for calculating, for each image area, a feature amount that is a difference between the images from a plurality of images having different shooting times;
Detecting means for detecting an image area having a change in the feature amount based on the feature amount of each image area calculated by the feature amount calculating unit;
A specifying means for specifying an image area to be excluded from object detection as an excluded area from the plurality of image areas;
Output means for outputting information on an image area obtained by removing the exclusion area specified by the specifying means from the image area detected by the detecting means;
Determination means for determining whether an object exists in the imaging range that is the subject based on the information output by the output means;
With
The detection unit is a detection device that specifies an image area to be excluded, which is associated with time or time , as the excluded area when excluded by the exclusion setting information . A computer of an information output device that outputs information about an image obtained by photographing a subject,
Dividing means for dividing the image into a plurality of image areas;
Feature amount calculating means for calculating, for each image area, a feature amount that is a difference between the images from a plurality of images having different shooting times;
Detecting means for detecting an image area having a change in the feature amount based on the feature amount of each image area calculated by the feature amount calculating unit;
A specifying means for specifying an image area to be excluded from information output as an excluded area from the plurality of image areas;
Output means for outputting information on an image area obtained by removing the exclusion area specified by the specifying means from the image area detected by the detecting means;
To function,
The specifying unit is a program for specifying, as the excluded area , an image area to be excluded, which is associated with time or time when excluded by exclusion setting information .   The program according to claim 8, wherein the specifying unit specifies one or a plurality of image areas as a processing unit and specifies the exclusion area for each processing unit. A method executed by an information output device that outputs information about an image obtained by photographing a subject,
A dividing step of dividing the image into a plurality of image areas;
A feature amount calculating step for calculating, for each image area, a feature amount that is a difference between the images from a plurality of images having different shooting times;
A detecting step for detecting an image area having a change in the feature amount based on the feature amount of each image area calculated by the feature amount calculating unit;
A specifying step of specifying, from the plurality of image areas, an image area to be excluded from information output as an excluded area;
An output step for outputting information relating to an image area obtained by removing the exclusion area specified by the specifying unit from the image area detected by the detecting unit;
Including
The information output method for specifying the image area to be excluded, which is associated with time or time , as the exclusion area when the identification means is excluded by the exclusion setting information .   The information output method according to claim 10, wherein the specifying unit specifies one or a plurality of image areas as a processing unit and specifies the exclusion area for each processing unit.

Images