JP4512763B2 - Image shooting system - Google Patents

Description

  The present invention relates to an image photographing system for photographing an image while protecting the privacy of a photographing target.

  With the development of digital video equipment and the widespread use of mobile phones with camera functions, it is becoming possible for anyone to take pictures and images anytime and anywhere. Research on intelligent video media aimed at recognizing and understanding various events and presenting information according to the situation and demands of observers by integrating a large amount of video information acquired by such sensors. It is actively carried out (for example, see Non-Patent Document 1).

  These studies are expected to realize smooth communication between users across time and space and the sharing of surrounding environmental information. However, if a system that captures, records, transmits, and presents events without considering privacy concerns such as infringement of portrait rights on subjects and restrictions on free behavior, the system is very inconvenient for users. It is thought that it becomes. In particular, it is not uncommon for lawsuits to be filed for shooting and distributing videos without subject permission.

  In order to solve such a problem, a technique is used in which privacy is protected by performing image processing such as blurring on an area that may infringe privacy in a video and reducing the amount of visual information. .

  For example, in the UK, millions of surveillance cameras have been placed in the city, and there is a track record of actually solving crimes using those images. However, areas such as windows of people's houses in the images are called “private zones”. It is called, and the privacy problem is avoided by applying the black painting process.

  The camera used in such a surveillance video shooting / presentation system is fixed in the shooting space, and the area where privacy should be protected is stationary. For this reason, the area on which the image processing is performed on the photographed video does not change, but the area on which the image processing is performed also changes dynamically when the photographing camera or the subject whose privacy is to be protected moves.

  However, there is a problem that it is not realistic to deal with this area change manually in an intelligent video media that handles a large amount of video data.

In addition, as another technique for capturing an image for privacy protection, there is a technique for detecting the body and face of a subject on the captured image and blurring the area. In this case, detection processing is often performed using a result obtained by dividing a captured image into a foreground region and a background region (see, for example, Non-Patent Document 2). However, it is difficult to divide an area using a background difference process or an inter-frame difference process for a video shot while moving the camera.
A. Pentland, “Looking at People: Sensing for Ubiquitous and Wearable Computing”, IEEE Trans. On Pattern Analysis and Machine Intelligence, Vol. 22 (1), pp. 107-118, 2000. H. Murase, SKNayer, “Parametric Eigenspace Representation for Visual Learning and Recognition”, Workshop on Geometric Method in Computer Vision, SPIE, pp.378-391, 1993.

  Therefore, even in applications such as surveillance cameras, privacy protection in real time is possible even when the object (subject) that protects privacy moves in the area to be photographed or when the camera for photographing itself moves. For this reason, it is desirable to dynamically change the area to be subjected to image processing.

  On the other hand, if priority is given to privacy protection, as described above, for the subject whose privacy is to be protected, the “blurring process” or the like may be performed on the image information portion that can identify the subject. Therefore, there may be a request to reproduce the image information portion of the area subjected to the blurring process. For example, if a surveillance camera later knows that an image related to a crime has been taken, it wants to reproduce the image information part of the area subjected to the blurring process by releasing the blurring process. This is the case when there is a demand.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to selectively perform image processing, such as blurring processing, on an area that needs privacy protection in a captured image. It is an object to provide an image photographing system capable of performing the above.

  Another object of the present invention is that, for privacy protection, an area where image processing such as blurring processing is usually performed, a user having a predetermined authority views an image for which the image processing has been canceled afterwards. It is an object of the present invention to provide an image photographing system capable of doing this.

In order to achieve such an object, the image photographing apparatus system of the present invention includes three-dimensional position detecting means for detecting the positions of a plurality of moving subjects to be photographed in a three-dimensional space where photographing is performed. In the three-dimensional space, a plurality of LED tags whose positions are specified in advance are arranged, and in a state of moving in the three-dimensional space, a moving photographing means for photographing an image of the photographing subject is photographed. A protected area candidate detecting means for detecting a face area as a candidate area in a photographed person image that is an object of image processing for protecting privacy, and privacy protection targets among the detected candidate areas. and a concealing portion detecting means for detecting a composed protected area, concealing portion detecting means, candidate territory corresponding to the photographic subject that is not with the identification mark of the plurality of imaging subject It detects a protected area, the protective area identified by concealing part detecting means, selectively and a protective image generating means for generating a protective image data subjected to image processing for protection of the privacy The protection image generating means calibrates the moving photographing means based on the ID information sent by blinking of the LED tag present in the image photographed by the moving photographing means to derive a projection matrix for the moving photographing means. , the protection region is identified by calculating how projected on any portion in the captured image, that perform image processing.

  Preferably, the concealed portion detection means includes a protection area separation means for selectively extracting image data corresponding to the protection area, and encrypts and encrypts the image data corresponding to the protection area extracted by the protection area separation means. It further includes a protected area encryption means for generating image data.

  Preferably, an image compression unit for compressing the protected image data that is output from the protected image generation unit to generate compressed image data, the compressed image data, and the encrypted image data are received, and the corresponding image is reproduced. Image browsing means for authenticating the user, and the image browsing means reproduces an image obtained by decompressing the compressed image data and an image obtained by decrypting the encrypted image data for the authenticated user. Image reconstruction means for reproducing and outputting the composed image is included.

  Preferably, the protected area encryption unit encrypts the image data corresponding to the protected area without compression.

  Preferably, the protected image data and the encrypted image data are output as separate files.

  Preferably, the image processing apparatus further includes output image generation means for integrating the protected image data and the encrypted image data and outputting the integrated image data as one file.

  In the image photographing system of the present invention, it is possible to dynamically change the area where image processing for privacy protection is performed in real time. Furthermore, according to the present invention, only a user who has been subjected to predetermined authentication can view an image portion that has been subjected to image processing such as blurring processing after canceling the image processing.

(Basic configuration and operation of the invention)
Embodiments of the present invention will be described below with reference to the drawings.

  The present invention provides a video acquisition method for realizing recognition and understanding of an event and presentation of information according to an observer's situation / request while protecting privacy even for dynamically changing video input. . Hereinafter, such a system is referred to as a stealth vision system. In the stealth vision system of the present invention, an area in which privacy is to be protected (privacy protection area) is set as a three-dimensional model, and is projected onto a video photographed by a calibrated mobile camera or the like. An area for image processing on a moving object is appropriately set.

[System configuration of the present invention]
FIG. 1 is a conceptual diagram showing an example of a stealth vision system 1000 using the image photographing system of the present invention.

  With reference to FIG. 1, the system 1000 will be described later based on a ceiling camera 10 for taking a bird's-eye view of a three-dimensional space such as a room where photographing is performed, and an image from the ceiling camera 10. In addition, the object to be photographed, for example, the subjects 4 and 6 is tracked, and an image processing such as blurring (including mosaic processing) is performed in the three-dimensional space for privacy protection (hereinafter referred to as “privacy protection region”). And a communication device 200 that outputs a signal for specifying the privacy protection area from the computer 100. Although not particularly limited, for example, the communication device 200 can be a wireless device.

  The system 1000 further captures the moving camera 20 held by the photographer 2 to photograph the subjects 4 and 6, and the photographing signal from the moving camera 20, and records the privacy-protected image data on the recording medium. Privacy protection mobile video generation device 300. For example, a mobile computer capable of recording data on a large-capacity recording medium can be used as the privacy protection mobile video generation apparatus 300. In FIG. 1, the computer 100 and the privacy protection mobile video generation device 300 are described as separate bodies, but the video signal from the ceiling camera 10 is transmitted, for example, wirelessly, and the privacy protection mobile video generation device The computer 100 and the privacy-preserving mobile video generation device 300 can be a wearable device for the photographer 2 if the device 300 can function as the above-described three-dimensional position detection unit. It is.

  As will be described later, the system 1000 further includes an LED-ID (Light Emitting Diode-ID) tag 30 arranged in the three-dimensional space, and the subjects 4 and 6 to calibrate the camera. In order to identify the subject 6 to be subjected to image processing for privacy protection, a tag 40 attached to the body of each of the subjects 4 and 6 is provided. As the tag 40, for example, an RFID (Radio Frequency-Identification) tag can be used. The computer 100 can specify the position of the RF tag in the three-dimensional space by the communication device 200. In this case, the position specifying accuracy by the RF tag does not need to be as high as the position accuracy obtained from the image captured by the ceiling camera 10.

  In the following description, the camera 20 is described as a moving camera, but the camera 20 itself may be a fixed camera for photographing a subject.

  FIG. 2 is a block diagram showing the configuration of the computer 100.

  Referring to FIG. 2, this computer 100 reads and writes information to and from an optical drive 108 and a flexible disk (FD) 116 for reading information on a disk medium such as a CD-ROM (Compact Disc Read-Only Memory). A computer main body 102 having an FD drive 106 for displaying the image, a display 103 as a display device connected to the computer main body 102, and a keyboard 110 and a mouse 112 as input devices also connected to the computer main body 102.

  Further, as shown in FIG. 2, the computer main body 102 constituting the computer 100 includes a central processing unit (CPU) 120 and a ROM (ROM) connected to the bus BS in addition to the optical drive 108 and the FD drive 106. A memory 122 including a read only memory (RAM) and a random access memory (RAM), a direct access memory device, for example, a hard disk 124, and a communication interface 128 for exchanging data with the communication device 200 are included. For example, a CD-ROM 118 is attached to the optical drive 108. An FD 116 is attached to the FD drive 106.

  The CD-ROM 118 may be another medium, such as a DVD-ROM (Digital Versatile Disc) or a memory card, as long as it can record information such as a program installed in the computer main body. In this case, the computer main body 102 is provided with a drive device that can read these media.

  Note that the privacy protection mobile video generation apparatus 300 is basically a mobile computer having the same configuration as the computer 100 by using the hard disk 124 or other large-capacity recording medium as a recording medium for image data. Can be realized. Of course, the privacy-preserving mobile video generation device 300 may be configured with dedicated hardware instead of a general-purpose mobile computer.

[Stealth vision system using mobile camera]
The stealth vision system 1000 captures an unspecified moving person / object with the moving camera 20, and also adds problems such as changes in illumination conditions and hiding between objects, so detection using appearance information on video This is generally difficult to achieve with techniques.

  Therefore, in order to solve this problem, as shown in FIG. 1, in the present invention, a privacy protection region is set in a three-dimensional space using a plurality of cameras. In the present invention, since the estimated depth information of the subject is used, it is possible to set an appropriate blur region even when an overlap between objects occurs on the captured image. In addition, since this setting process does not use any appearance information on the video, there is an advantage that it is not affected by changes in appearance due to movement of objects, changes in posture, and illumination conditions.

  FIG. 3 is a functional block diagram for showing an overview of the configuration and operation of the system 1000 shown in FIG.

  As described above, the system 1000 includes the three-dimensional position detection unit and the privacy protection mobile video generation apparatus 300.

  The three-dimensional position detection unit performs three-dimensional modeling of the photographing space using at least one stationary camera attached to the three-dimensional space, for example, the ceiling camera 10. In the modeled three-dimensional area, if video shooting is not permitted for privacy protection, that area is set as a privacy protection area.

  In the privacy protection mobile video generation device 300, the mobile camera 20 is calibrated, the mobile image capture unit 302 acquires an image from the mobile camera 20, and the privacy protection image generation unit 304 captures the image plane as a three-dimensional space. Estimate the projective relationship between. Using this projection relationship, where the privacy protection area is reflected on the image is obtained, and image processing is performed on the estimation area, thereby realizing privacy protection. Privacy protected image data generated by the privacy protected image generation unit 304 is recorded on a recording medium.

[Privacy-protected video generation method for the subject]
(Privacy protection area setting)
When restoring the three-dimensional shape of an object to be photographed, the problem of occlusion between objects can be solved by applying stereo processing or the like to images photographed by a large number of cameras, but the calculation processing cost is high. There is a problem of growing.

  In the present invention, in order to realize a stealth vision system that operates in real time, not a large number of cameras but a single camera (ceiling camera 10) that captures the space shown in FIG. The privacy protection area is estimated.

  Since the imaging target of the stealth vision system is an object that moves almost upright on the ground, there is an advantage that the problem of hiding between objects hardly occurs in an image from above.

  FIG. 4 is a conceptual diagram showing a procedure for estimating the privacy protection area from the photographed image of the ceiling camera 10.

  In order to estimate the three-dimensional position (X, Y, Z) from the two-dimensional coordinate information (u, v) on the ceiling camera image, it is assumed that all objects exist on a plane having a height Y. If the object to be photographed is a human who moves in a three-dimensional space, this assumption sufficiently reflects reality.

  From the two-dimensional projective transformation matrix H between the plane and the ceiling camera image plane, the position in the three-dimensional space is estimated from the two-dimensional coordinate information on the image as shown in the following equation (1).

λ [X Z 1] ^T = H _n [u v 1] ^T (1)
For such estimation methods, for example, Reference 1: T. Koyama, I. Kitahara, Y. Ohta, “Live Mixed-Reality 3D Video in Soccer Stadium”, Proc. Of IEEE and ACM Int. Symposium on Mixed and Augmented Reality (ISMAR2003), pp178-187, (2003).

  In the processing using Expression (1), the two-dimensional coordinates of the object can be evaluated by simple image processing such as background removal and foreground area labeling. For this reason, such processing is suitable for real-time processing. In order to reduce the calculation load, the privacy protection area is assumed to be a box-like area surrounding at least a part of the object (for example, the face of the subject). It is assumed that the size of the box-shaped region is appropriately changed for each photographing target. Therefore, when acquiring the privacy protection area of the face of the subject, it is assumed that the subject exists in an area having a height of 1.5 m from the ground in Expression (1), and the calculated three-dimensional position is obtained as a result. This can be realized by setting a box-like region of 30 cm cubic around the area.

  In this embodiment, since the focus is on real-time processing, a box-like shape with a small amount of calculation is adopted as the privacy protection area, but the privacy protection area set in the present invention is not limited to this. is not. If the calculation cost permits, it is possible to set an accurate three-dimensional shape of the privacy protection region by applying a three-dimensional shape restoration process to an image captured by a multi-view camera. By using this shape, it can be expected to improve the accuracy of the subsequent blurring process.

(Calibration of moving camera)
Hereinafter, a procedure for calibrating the mobile camera 20 will be briefly described.

  The calibration of the camera is disclosed in detail in, for example, Document 2: Xu Tsuyoshi and Saburo Tsubaki “3D Vision”, Kyoritsu Shuppan, April 20, 1998, first edition.

  As a method for calibrating the moving camera, first, there is a method in which external variables of the camera such as the position and direction of the moving camera are extracted by a group of three-dimensional position sensors using magnetic / infrared rays. The projection matrix is obtained by combining the camera internal variables and external variables.

  As a second method, the projection matrix is evaluated by a large number of pairs of three-dimensional coordinates and two-dimensional coordinates.

  In the first method, although a projection matrix can be obtained stably, the observable region is limited due to the number of available reception sensors.

  In the second method, since the moving camera itself is a reception sensor, the easily observable area can be expanded by increasing the number of signs arranged in the three-dimensional space. However, as a practical problem, it is difficult to accurately extract the two-dimensional coordinates of the sign, and it is generally difficult to stably obtain a projection matrix.

  In the present invention, as shown in FIG. 1, the LED-ID tag 30 is arranged as a marker in the three-dimensional space. For example, the LED-ID tag 30 can be created with a bottom surface size of about 3 cm × 4 cm, so that it can be easily placed anywhere in the three-dimensional space where photographing is performed.

  Each LED-ID tag 30 transmits its own ID number in a blinking pattern, and thus it is possible to specify which tag the LED-ID tag 30 shown in the image is. . The internal variables of the camera are obtained in advance, and the LED-ID tag 30 is detected in the captured image, and the ID number is decoded and specified. With this ID number, the three-dimensional coordinates of the LED-ID tag 30 obtained in advance can be referred to. It is assumed that such an ID number and three-dimensional coordinates are stored in advance in the storage device of the computer 100.

  If there are at least three LED-ID tags 30 in the captured image, an external variable of the camera can be obtained from these three-dimensional coordinates. A projection matrix can be obtained by combining external variables and internal variables obtained in this way.

(Operation of system 1000)
Hereinafter, the operation of the computer 100 that operates as a three-dimensional position detection unit and the operation of the privacy protection mobile video generation apparatus 300 will be described.

  FIG. 5 is a flowchart for explaining the operation of the computer 100.

  Referring to FIG. 5, first, the ceiling camera 10 is calibrated (step S100). In this case, for example, when applying the expression (1), if it is assumed that the privacy protection area of the subject is at a height of 1.5 m from the ground, a plurality of calibration bars having a height of 1.5 m are provided. They are arranged vertically at predetermined intervals in a three-dimensional space. A colored marker is attached to the tip of the calibration bar, and the ceiling camera 10 can be calibrated with this marker.

  Subsequently, based on the RFID tag 40 and the image data from the ceiling camera 10, it is detected whether the privacy protection target exists in the three-dimensional space to be photographed (step S102).

  If there is a privacy protection object, a privacy protection area (box-shaped area) in the three-dimensional space is calculated according to the equation (1) (step S106). Subsequently, information about the calculated privacy protection area is notified to the privacy protection image generation unit 304 of the privacy protection mobile video generation device 300 via the communication device 200 (step S108). When it is determined in step S104 that no privacy protection target exists, information indicating that is notified.

  If the operator has not instructed the end of the process, the process returns to step S102 again.

  FIG. 6 is a flowchart for explaining the operation of the privacy protection mobile video generation apparatus 300.

  Referring to FIG. 6, first, a video signal is captured by mobile camera 20 (step S200). Subsequently, the mobile camera is calibrated using the LED-ID tag 30 described above (step S202).

  If the privacy protection area has been notified (step S204), a blurring process is performed on a portion corresponding to the privacy protection area in the two-dimensional image (step S206).

  Here, using the box-shaped privacy protection region estimated in the step of calculating the privacy protection region as described above and the projection transformation matrix P calculated by the calibration process of the mobile camera, A procedure for performing image processing (such as blurring processing) for protecting privacy will be described.

  As described above, at least a part of the subject in the target space is expressed as a box-shaped region. Eight vertices of the area are projected on an image photographed by the moving camera according to the following expression (2), and a convex area Rn connecting these points is set as an observation area of the subject on the screen.

λ [u _n v _n 1] ^T = P [X _n Y _n Z _n 1] ^T (2)
(N = 1, ..., 8)
In this observation area, blurring processing for reducing the resolution of the image, processing for painting with a uniform color, and the like are performed, so that the appearance information of the subject is reduced and privacy protection is realized.

  The image data after the blurring process is performed, or if the privacy protection area is not notified, the captured image data is output as image data and recorded on the recording medium (step S208).

  If the operator has not instructed the end of the process, the process returns to step S200 again.

  FIG. 7 is a diagram showing an example of an image subjected to privacy protection processing as described above.

  As shown in FIG. 7, a mosaic process is performed on the face area of the person from the center of the screen in accordance with the change in the size of the face part of the person.

(Embodiment 1)
In the above description, it has been described that the blur processing is performed on the face area of the person to whom the tag (RFID tag) 40 is attached. However, if privacy protection is required for an actual application scene, for example, an image taken by a surveillance camera in a predetermined facility, rather than an employee working in the facility, that is, on the photographed image. A subject that does not necessarily require privacy protection has a tag 40, but a visitor or the like of the facility often does not have the tag 40 attached.

  In this case, in the process of step S102 in FIG. 5, first, an image region portion corresponding to the face is detected by a known method in the image taken by the camera 20, and image information from such a camera 20 is detected. Based on the detection result of the face based on the information, the information from the tag 40, and the image data from the ceiling camera 10, it is detected whether the privacy protection target exists in the three-dimensional space to be photographed. That is, it is determined that a privacy protection area exists for a subject that does not have the tag 40 among the faces detected in the image of the camera 20.

  Below, the modification which performs a blurring process about the subject which does not have such a tag 40 is demonstrated. In addition, although not particularly limited, in the following, the camera 20 is assumed to be fixed and photographed.

  FIG. 8 is a schematic block diagram for explaining the configuration of an image processing apparatus 306 for performing privacy protection processing on an image captured in this way.

  It is assumed that the subject 6 of the subjects 4 and 6 has the tag 40 attached.

  In FIG. 8, the camera 20 captures an image including a subject, and the tag detection unit 22 detects whether the tag 40 exists within the capturing range. It is also assumed that a ceiling camera 10 (not shown) is also installed.

  Although not particularly limited, each function of the image processing apparatus 306 can also be realized by software executed on a computer.

  In the image processing device 306, first, the face area detection unit 310 detects a face area in the photographed image, and the concealed part selection unit 312 detects information from the tag detection unit 22 and a ceiling camera as necessary. Based on information from a three-dimensional position detection unit (not shown) that processes image information from 10, a portion (privacy protection area) for performing blurring processing is specified in the photographed image.

  Subsequently, the face area concealed image generation unit 314 calculates a privacy protection area (for example, a box-shaped area) in the three-dimensional space for the identified privacy protection area, as described above. Blur processing is performed on a portion corresponding to the privacy protection area in the two-dimensional image. The converted image output unit 318 outputs an image subjected to the blurring process.

  With the configuration as described above, blurring processing for privacy protection can be performed on a subject without the tag 40. As a result, in order to protect the privacy information that is collected outside the original purpose, the necessary information that is the purpose (of the people who are to collect information or those who are allowed to collect information) It is possible to conceal only the part that causes a privacy problem (such as a face image other than the information collection target person) while leaving the image).

  Note that an LED-ID tag may be used as the tag 40 instead of the RFID tag. In this case, whether or not the subject is wearing the LED-ID tag based on image information captured by the camera 20 is determined. Can be detected.

(Embodiment 2)
In the first embodiment, the configuration for simply generating and outputting an image subjected to the blurring process for the privacy protection area has been described.

  However, as described above, in some cases, an image that has been subjected to the blurring process for the privacy protection area may be later viewed after the blurring process that has been performed on the privacy protection area is canceled.

  Therefore, in the second embodiment, an image photographing system capable of such browsing will be described.

  FIG. 9 is a schematic block diagram for explaining the configuration of the image processing apparatus 306 used in the image photographing system of the second embodiment.

  In FIG. 9, the same parts as those in FIG. In FIG. 9, only the subject in which the privacy protection area exists is shown for ease of explanation.

  In the figure, it is assumed that the camera 20 can capture a high-resolution image (for example, an image with a high-definition image quality).

  Referring to FIG. 9, an image of subject 4 is captured by camera 20, and a face is detected by a well-known method in face area detection unit 310 of image processing device 306. On the other hand, based on information from the tag detection unit 22 and information from a ceiling camera (not shown), the image information of the face area of the subject where the privacy protection area exists in the photographing area is selectively selected. Extracted in

  For the image after the extraction processing by the face region separation unit 313 is finished, the face region concealed image generation unit 314 performs privacy protection in the three-dimensional space for the identified privacy protection region as described above. A region (for example, a box-shaped region) is calculated, and a blurring process is performed on a portion corresponding to the privacy protection region in the two-dimensional image. The image processed by the face area hidden image generation unit 314 is subjected to compression processing by the image compression unit 316, and then the image subjected to blurring processing is output from the converted image output unit 318.

  On the other hand, the image corresponding to the privacy protection area extracted by the face area separation unit 313 is subjected to encryption processing by the face area image encryption unit 320 using the encryption key generated by the encryption key generation unit 322. The Although not particularly limited, the encryption key generated by the encryption key generation unit 322 may be generated for each shooting operation (from camera on to off) or updated at predetermined time intervals. It may be done. In any case, the information output from the face area image encryption unit 320 includes the encrypted face area image information and information that can specify the encryption key that encrypted the face area image information. Shall be included. The output of the face area image encryption unit 320 is stored in the encryption information storage unit 324.

  FIG. 10 shows a compressed image file (compressed image with a concealed face image) output from the converted image output unit 318 and an encrypted file (encrypted face information) stored in the encryption information storage unit 324. It is a conceptual diagram showing typically.

  It is assumed that the compressed image file and the encrypted file are associated with each other by, for example, time information when the image is taken.

  FIG. 11 is a schematic block diagram for explaining the configuration of an image browsing apparatus 400 used in the image capturing system of the second embodiment.

  Referring to FIG. 11, image browsing apparatus 400 receives a compressed image file data from converted image output unit 318, and decompresses the compressed image file data from converted image input unit 410. A compressed image expansion unit 420.

  The image browsing apparatus 400 further receives, from the encryption information storage unit 324, an encryption information input unit 412 that receives encrypted file data when the user is authenticated, and an encryption key from the encryption key generation unit 322. A face area image decoding unit 414 for decrypting the face area image information encrypted by the encryption key when the user is authenticated, and a compressed image expansion unit 420 when the user is not authenticated. If the user authentication is performed, the information from the compressed image decompression unit 420 and the decoded information from the face image decoding unit 414 are integrated to reconstruct the image. The image reconstruction unit 424, the authentication information input unit 426 that receives authentication information from the user and determines whether or not to authenticate the user, and the output from the image reconstruction unit 424 And an image output unit 430 for and view.

  In the image collecting apparatus that automatically extracts a person in a captured image and hides only that portion by using the configuration as shown in FIGS. (Privacy protection area, for example, the face part) is separated, and the image of the part having the privacy problem is stored as information associated with the concealed image different from the concealed image. At times (when the original image is necessary due to the occurrence of a crime, etc.), the original image can be restored to the user who has received the predetermined authentication.

  In the configuration shown in FIG. 9, as in the case shown in FIG. 8, the person who is shown in the image is recognized and the person identified as having no privacy problem is identified. Thus, it is possible to perform an operation of hiding only a region having privacy problems of other people without processing an image of a region having privacy problems.

  Furthermore, the face area image is encrypted without compressing the captured high-quality image data, and the image of the area having a privacy problem is set as a higher resolution image than the other parts. Since it is saved, it becomes easy for a user who has received predetermined authentication to specify a person who has moved to an area concealed as a privacy protection area when the original image is restored.

(Embodiment 3)
In the second embodiment, the data obtained by encrypting the privacy protection area is stored as a separate file from the compressed image data.

  However, the compressed image data and the data obtained by encrypting the privacy protection area can be stored in different areas of the same file.

  For example, when a moving image is encoded by an image encoding method such as motion JPEG that encodes every frame, the compressed image data is stored as the original JPEG compressed image data body, and the privacy is stored. The encrypted data in the protected area can be stored in an area that is not normally reproduced during image reproduction, such as a comment area. In addition, even in the case of an image encoding method that compresses in the time axis direction, basically, the encrypted data in the privacy protection area is usually reproduced for each frame in a format that conforms to the encoding method. Store it in an area that is sometimes not played back.

  FIG. 12 is a schematic block diagram for explaining the configuration of the image processing apparatus 306 used in the image photographing system of the third embodiment.

  In FIG. 12, the same parts as those in FIG. Hereinafter, differences from the configuration of FIG. 9 will be mainly described.

  In the image processing apparatus 306 illustrated in FIG. 12, the image corresponding to the privacy protection area extracted by the face area separation unit 313 is subjected to face area image encryption using the encryption key generated by the encryption key generation unit 322. The encrypted data subjected to the encryption processing in the unit 320 is given to the output image generation unit 317 together with the compressed image data generated by the image compression unit 316.

  The output image generation unit 317 integrates the compressed image data and the encrypted data into one compressed image file, and provides the converted image output unit 318 with it. However, in the compressed image file, the encrypted face information stored in the comment part is not compressed. Since the image area of the face image portion is not originally large, the data size itself does not increase even in this way.

  Since other configurations and operations are the same as those shown in FIG. 9, the description thereof will not be repeated.

  FIG. 13 is a conceptual diagram illustrating a configuration of a compressed image file generated by the output image generation unit 317.

  The main part (data part) of one image file stores the compressed image data concealed in the face image, and is encrypted in an area that is not normally reproduced during image reproduction like the comment part of the file. Face information is stored.

  FIG. 14 is a schematic block diagram for explaining the configuration of the image browsing device 406 used in the image capturing system of the third embodiment.

  Also in FIG. 14, the same parts as those in FIG. Hereinafter, differences from the configuration of FIG. 11 will be mainly described.

  Referring to FIG. 14, the converted image input unit 410 separates the input compressed image data into a data part and a comment part, and the compressed image data is a compressed image decompression unit for decompressing the compressed image data. The encrypted data is provided to the face area image decrypting unit 414.

  Subsequent processing is the same as the operation of the image browsing apparatus 400 of FIG.

  Even with the configuration as described above, it is possible to achieve the same effect as the image capturing system of the second embodiment.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is a conceptual diagram which shows an example of the stealth vision system 1000 using the image imaging device of this invention. FIG. 2 is a block diagram showing the configuration of a computer 100. It is a functional block diagram for showing the outline | summary of a structure and operation | movement of the system 1000 shown in FIG. It is a conceptual diagram which shows the procedure which estimates a privacy protection area | region from the picked-up image of the ceiling camera. 4 is a flowchart for explaining the operation of the computer 100. 5 is a flowchart for explaining an operation of a privacy-protected mobile video generation device 300. It is a figure which shows the example of the image by which privacy protection processing was carried out. It is a schematic block diagram for demonstrating the structure of the image processing apparatus 306 for performing a privacy protection process. FIG. 6 is a schematic block diagram for explaining a configuration of an image processing device 306 used in the image capturing system of the second embodiment. 4 is a conceptual diagram schematically illustrating a compressed image file output from a converted image output unit 318 and an encrypted file stored in an encryption information storage unit 324. FIG. It is a schematic block diagram for demonstrating the structure of the image browsing apparatus 400 used in the image imaging system of Embodiment 2. FIG. FIG. 10 is a schematic block diagram for explaining a configuration of an image processing device 306 used in the image capturing system of the third embodiment. 3 is a conceptual diagram illustrating a configuration of a compressed image file generated by an output image generation unit 317. FIG. FIG. 10 is a schematic block diagram for explaining a configuration of an image browsing device 406 used in the image capturing system of the third embodiment.

Explanation of symbols

  100 computer, 102 computer main body, 103 display, 106 FD drive, 108 optical drive, 110 keyboard, 112 mouse, 114 wireless communication device, 118 CD-ROM, 120 CPU, 122 memory, 124 hard disk, 128 communication interface, 200 communication device , 300 Privacy protection mobile video generation device, 1000 image capturing system.

Claims (6)

A three-dimensional position detecting unit is provided for detecting the positions of a plurality of moving subjects in the three-dimensional space where photographing is performed. In the three-dimensional space, a plurality of LED tags whose positions are specified in advance are provided. Has been placed,
In a state of moving in the three-dimensional space, a moving photographing means for photographing an image of the photographing subject;
Protected area candidate detecting means for detecting a face area as a candidate area in a photographed person image to be subjected to image processing for protecting privacy in a captured image;
A concealed part detecting means for detecting a protected area that is subject to privacy protection among the detected candidate areas, and the concealed part detecting means attaches an identification mark among the plurality of subjects to be photographed. Detecting the candidate area corresponding to a non-shooting target person as the protection area,
Wherein with respect to the protected area specified by the concealment portion detecting unit, selectively a protective image generating means for generating a protective image data the image processing was carried out to protect privacy,
The protected image generating means calibrates the moving photographing means based on the ID information sent out by blinking of the LED tag existing in the image photographed by the moving photographing means. derive a projection matrix, the protected area is identified by calculating how is projected to any part in the photographed image, that perform the image processing, the image capturing system. The concealed part detecting means includes a protected area separating means for selectively extracting image data corresponding to the protected area,
The image photographing system according to claim 1, further comprising a protection area encryption unit that encrypts image data corresponding to the protection area extracted by the protection area separation unit to generate encrypted image data. Image compression means for generating compressed image data by compressing the protected image data that is the output of the protected image generation means;
An image browsing means for receiving the compressed image data and the encrypted image data and reproducing the corresponding image;
The image browsing means includes:
Image reconstructing means for authenticating a user and reproducing and outputting an image obtained by reconstructing an image obtained by decompressing the compressed image data and an image obtained by decrypting the encrypted image data for the authenticated user The image capturing system according to claim 2, comprising:   The image capturing system according to claim 3, wherein the protected area encryption unit encrypts image data corresponding to the protected area without compression.   The image capturing system according to claim 2, wherein the protected image data and the encrypted image data are output as separate files.   The image capturing system according to claim 2, further comprising output image generation means for integrating the protected image data and the encrypted image data and outputting the integrated image data as one file.