公开(公告)号：CA2420068C

公开(公告)日：2010-05-25

申请号：CA2420068

申请日：2002-06-26

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  NAGANO TAKAHIRO ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  FUJIWARA NAOKI ,  WADA SEIJI ,  MIYAKE TORU

IPC: G06T3/00 ,  G06T5/00 ,  G06F17/00 ,  G06T1/00 ,  G06T5/50 ,  G06T7/20 ,  H04N5/765 ,  H04N5/92 ,  H04N7/173 ,  H04N7/18 ,  H04N21/234 ,  H04N21/27 ,  H04N21/4223 ,  H04N21/431 ,  H04N21/44

Abstract: A picked up image is separated into a foreground component image and a background component image in real time. An image picked up by a pickup block (74) is separated into a foreground component image and a background component image which are accumulated in an image accumulation block (72) and a charging block (75) performs charging required for the separation. Here, a separation block (91) performs a motion unsharpness remedy to the separated foreground component image and outputs it together with the background component image to a synthesis block (92). The synthesis block (92) synthesizes the foreground component image which has been subjected to the motion unsharpness remedy and input with the separated background component image so as to create a synthesized image and displays it on a display block (73). The present invention can be applied to a camera terminal apparatus.

公开(公告)号：CA2420827C

公开(公告)日：2010-03-23

申请号：CA2420827

申请日：2002-06-25

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  FUJIWARA NAOKI ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  NAGANO TAKAHIRO ,  MIYAKE TORU ,  WADA SEIJI

IPC: G06T7/20 ,  G06T1/00 ,  G06T3/00 ,  G06T5/00 ,  H04N5/232 ,  H04N5/262

Abstract: An object of the present invention is to enable the detection of a mixture ratio indicating the mixture state of a plurality of objects, such as a background image and a moving object image. A normal-equation adder 541 extracts mixed pixel data according to a motion, and also extracts background pixel data corresponding to the mixed pixel data so as to generate relational expressions for a designated pixel. A normal-equation calculator 542 detects the mixture ratio indicating the mixture state of the objects based on the relational expressions. The normal-equation adder 541 generates the plurality of relational expressions based on a first approximation in which foreground object components change substantially linearly, and a second approximation in which the mixture ratio changes substantially linearly. The present invention is applicable to a signal processing apparatus for processing images.

公开(公告)号：CA2420069C

公开(公告)日：2010-03-09

申请号：CA2420069

申请日：2002-06-26

Applicant: SONY CORP

Inventor： MIYAKE TORU ,  WADA SEIJI ,  SAWAO TAKASHI ,  FUJIWARA NAOKI ,  ISHIBASHI JUNICHI ,  KONDO TETSUJIRO ,  NAGANO TAKAHIRO

IPC: G06T3/00 ,  G06T7/20 ,  G06T1/00 ,  G06T5/00 ,  H04N5/262 ,  H04N19/00 ,  H04N19/132 ,  H04N19/137 ,  H04N19/167 ,  H04N19/196 ,  H04N19/20 ,  H04N19/423 ,  H04N19/50 ,  H04N19/51 ,  H04N19/543 ,  H04N19/59 ,  H04N19/85

Abstract: An object of the present invention is to make it possible to specify a background image area, a moving object image area, and an image area in which the mixture of the background image area and the moving object image area occurs. A mixture-ratio calculator 103 extracts pixel data of a peripheral frame corresponding to a designated pixel of a designated frame of the image data as background pixel data corresponding to a background object, and also extracts, designated pixel data of the designated pixel contained in the designated frame so as to generate a plurality of relational expressions indicating the relationship between the designated pixel data and the background pixel data. The mixture-ratio calculator 103 detects the mixture ratio indicating the mixture state of the objects based on the relational expressions. An area specifying unit 104 calculates a predictive error by substituting the detected mixture ratio into the relational expressions, and determines whether an area to which the designated pixel belongs is a covered background area, an uncovered background area, a foreground area, or a background area. The present invention is applicable to an image processing apparatus.

公开(公告)号：DE60225285T2

公开(公告)日：2009-03-05

申请号：DE60225285

申请日：2002-06-26

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  WADA SEIJI ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  FUJIWARA NAOKI ,  NAGANO TAKAHIRO ,  MIYAKE TORU

IPC: G06T3/00 ,  G06T7/20 ,  G06T1/00 ,  G06T5/00 ,  H04N5/262 ,  H04N5/272 ,  H04N19/00 ,  H04N19/51

Abstract: It is possible to know the mixture state in an image. With reference to image data, a region identification block (103) identifies a mixed region where a foreground object component constituting a foreground object is mixed with a background object component constituting a background object, and a non-mixed region consisting of a foreground region composed of only a foreground object and a background region composed of only a background object component.

公开(公告)号：DE60225138D1

公开(公告)日：2008-04-03

申请号：DE60225138

申请日：2002-06-25

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  FUJIWARA NAOKI ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  NAGANO TAKAHIRO ,  MIYAKE TORU ,  WADA SEIJI

IPC: G06T3/00 ,  G06T7/20 ,  G06T1/00 ,  G06T7/00 ,  H04N5/14 ,  H04N5/262

Abstract: The present invention allows the amount of movement indicating the movement of an object within an exposure period to be obtained. An estimated-mixture-ratio processor 401 calculates an estimated mixture ratio for each pixel and an estimated motion vector by a calculation corresponding to a model of a covered background area, and sends the calculated estimated mixture ratio and the calculated estimated motion vector to a mixture-ratio determining portion 403. An estimated-mixture-ratio processor 402 calculates an estimated mixture ratio for each pixel and an estimated motion vector by a calculation corresponding to a model of an uncovered background area, and sends the calculated estimated mixture ratio and the calculated estimated motion vector to the mixture-ratio determining portion 403. The mixture-ratio determining portion 403 sets a mixture ratio and a motion vector according to area information indicating to which of a foreground area, a background area, the covered background area, or the uncovered background area a designated pixel belongs. The present invention can be applied to image processing apparatuses.

公开(公告)号：CA2420719A1

公开(公告)日：2003-02-25

申请号：CA2420719

申请日：2002-06-20

Applicant: SONY CORP

Inventor： NAGANO TAKAHIRO ,  MIYAKE TORU ,  FUJIWARA NAOKI ,  KONDO TETSUJIRO ,  ISHIBASHI JUNICHI ,  WADA SEIJI ,  SAWAO TAKASHI

IPC: G06T1/00 ,  G06T5/00 ,  G06T7/20 ,  H04N5/272 ,  G06T3/00 ,  H04N5/262

Abstract: A background image can be separated from an object image in accordance with a mixture state. A region identification block (103) identifies at a first tim e, an uncovered background region where a foreground object component and a background object component are mixed, identifies at a second time, a foreground region composed of only foreground object components and backgrou nd region composed of only background object components, and identifies at a third time, a covered background region where the foreground object componen t and the background object component are mixed. A mixture ratio calculation block (104) detects a mixture ratio indicating a mixture ratio of the foreground object component and the background object component. A foregroun d and background separation block (105) separates the pixel data into a foreground object component and a background object component. The present invention can be applied to an image processing apparatus.

公开(公告)号：CA2420029A1

公开(公告)日：2003-02-18

申请号：CA2420029

申请日：2002-06-26

Applicant: SONY CORP

Inventor： MIYAKE TORU ,  FUJIWARA NAOKI ,  WADA SEIJI ,  KONDO TETSUJIRO ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  NAGANO TAKAHIRO

IPC: H04N5/262 ,  G06T7/20 ,  H04N5/00 ,  H04N7/16 ,  H04N7/173 ,  H04N7/18 ,  H04N7/24 ,  H04N7/36 ,  H04N7/50 ,  H04N19/00 ,  H04N19/20 ,  H04N19/25 ,  H04N19/46 ,  H04N19/51 ,  H04N19/70 ,  H04N21/232 ,  H04N21/2343 ,  H04N21/274 ,  H04N21/4223

Abstract: A foreground component image and a background component image of a picked up image are processed at a high speed by utilizing a network. A client compute r (27) outputs to a separation server (11) information specifying image data t o be separated. The separation server (11) fetches image data specified from a n accumulation server (18) and outputs it to a motion detection server (12). After the motion detection processing is performed, the separation server (1 1) outputs image data, a motion vector, and position information to an area identification server (13). The area identification server (13) generates ar ea information on the area of image data and outputs the area information in addition to the image data, the motion vector, and the position information to a mixture ratio calculation server (14). The mixture ratio calculation serve r (14) calculates a mixture ratio using the image data, the motion vector, the position information, and the region information, and according to these information, a foreground and background server (15) separates a foreground and a background of the input image. The present invention can be applied to an image processing business model.

公开(公告)号：CA2412304A1

公开(公告)日：2002-12-06

申请号：CA2412304

申请日：2002-04-01

Applicant: SONY CORP

Inventor： WADA SEIJI ,  MIYAKE TORU ,  NAGANO TAKAHIRO ,  FUJIWARA NAOKI ,  KONDO TETSUJIRO ,  SAWAO TAKASHI ,  ISHIBASHI JUNICHI

IPC: G06T5/20 ,  G06T3/00 ,  G06T5/00 ,  G06T7/20 ,  H04N5/21 ,  H04N5/232 ,  H04N5/262 ,  H04N5/272 ,  G06T1/00 ,  H04N5/335

Abstract: An image processing apparatus capable of removing a moving cloudiness contained in a defocused image. A region identification block (103) identifi es a non-mixed region including a foreground region consisting of a foreground object component constituting a foreground object and a background region consisting of a background object component constituting a background object , or a mixed region where a foreground object component is mixed with a background object component. In accordance with the region identification result and the like, an isolation/cloudiness removing block (1503) isolates a foreground object component and a background object component from pixel dat a of the mixed region and performs processing to remove moving cloudiness from the isolated foreground object component in a batch mode. This can take into consideration a difference between a signal detected by an image pickup element and a real world.

公开(公告)号：CA2365233A1

公开(公告)日：2001-07-05

申请号：CA2365233

申请日：2000-12-28

Applicant: SONY CORP

Inventor： MIYAKE TOHRU ,  NAGANO TAKAHIRO ,  KONDO TETSUJIRO ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  WADA SEIJI ,  FUJIWARA NAOKI

IPC: G06T1/00 ,  G06T5/00 ,  G06T5/50 ,  G06T7/20 ,  G06T3/00

Abstract: A signal processing unit (12) acquires a second signal obtained by detecting by a sensor a first signal (a real-world signal having a first dimension), having a second dimension lower than the first dimension, and containing strain with respect to the first signal, and generates a third signal reduce d in strain in comparison with the second signal by performing signal processi ng based on the second signal.

公开(公告)号：CA2419670C

公开(公告)日：2013-04-02

申请号：CA2419670

申请日：2002-06-19

Applicant: SONY CORP

Inventor： KONDO TETSUJIRO ,  ISHIBASHI JUNICHI ,  SAWAO TAKASHI ,  NAGANO TAKAHIRO ,  FUJIWARA NAOKI ,  MIYAKE TORU ,  WADA SEIJI

IPC: G06T3/00 ,  G06T7/20 ,  G06T1/00 ,  G06T5/00 ,  H04N5/14 ,  H04N5/262 ,  H04N5/272 ,  H04N19/00 ,  H04N19/51 ,  H04N19/513

Abstract: An apparatus for detecting a mixed area in frames. An object extracting unit extracts a foreground object from an input image and generates an area-specified object formed of the foreground object and a value indicating that the foreground objects belongs to a background area. A motion compensator compensates for the motion of the area-specified object based on a motion vector and positional information thereof. A subtracting unit subtracts the pixel value of a pixel belonging to a foreground object of a current frame from the corresponding pixel of the foreground object of a preceding frame, so as to obtain a frame difference between the pixels belonging to the foreground area. A threshold-value processor detects a mixed area based on the difference.