Image processing apparatus

• US 8,189,962 B2
• Filed: 12/17/2007
• Issued: 05/29/2012
• Est. Priority Date: 12/19/2006
• Status: Active Grant

• Current Assignee: Hitachi Kokusai Electric Incorporated (Hitachi, Ltd.)
• Original Assignee: Hitachi Kokusai Electric Incorporated (Hitachi, Ltd.)
• Inventors: Ito, Wataru, Ito, Mitsue
• Primary Examiner(s): ENTEZARI, MICHELLE M
• Application Number:
  US11/957,533
• Publication Number:
  US 20080166017A1
• Time in Patent Office: 1,625 Days
• Field of Search: None
• US Class Current: 382/293
• CPC Class Codes: G06V 2201/01 Solutions for problems related to non-uniform document background
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition
• CPC Class Codes: G08B 13/1968 Interfaces for setting up or customising the system

Abstract

In an image processing apparatus for processing images, a transformation unit transforms an image on a first coordinate system representing a coordinate system during pickup to an image on a second coordinate system set up on a reference plane, a display unit displays an image transformed by the transformation unit and a reception unit receives setting information a user inputs in accordance with a displayed image on the display unit. For example, in the image processing apparatus, the reception unit receives, as the setting information, information to be used for image processing.

Claims

1. An image processing apparatus for processing images picked up by a camera to detect objects in the images, comprising:

a memory for storing image data corresponding to the images picked up by the camera;

an image processor for processing said image data stored in said memory and detecting an object in said processed image data;

a display for displaying processed image data including the detected object output by said image processor; and

a designation unit for receiving setting information including coordinate transformation parameters and area parameters input by a user as a polygon or a circle defining an area in which a detected object is to be indicated to a user,

wherein said image processor transforms an image on a first coordinate system representing a coordinate system during pickup to an image on a second coordinate system set up on a plane representing a reference based on the coordinate transformation parameters, said display displaying the transformed image,

wherein said designation unit receives said polygon or the circle input by the user on the transformed image being displayed in the screen,

wherein said image processor transforms the picked-up image or/and the area parameters by using the coordinate transformation parameters and then detects an object inside the area defined by the area parameters for indicating to the user, and

wherein said second coordinate system is an approximation of an orthonormal coordinate system having two axes on a ground surface plane so that said image on the second coordinate system corresponds to a map of a monitoring target area, said coordinate transformation parameters includes three independent parameters corresponding to a depression angle, an installation height and a view field angle of said camera, and said image processor performs mutual transformation between the first coordinate system and the second coordinate system through central projection about the center at a principal point of the camera.

2. An image processing apparatus according to claim 1, wherein said setting information has size parameters for defining a range of actual height of the object, in normal direction to be detected.

3. An image processing apparatus according to claim 2, wherein each time that an object representing a reference is detected, said area parameter is updated to set up an area surrounding the reference object and said image processor decides whether an object other than said reference object is in said area under the second coordinate system, and then the object other than said reference object is indicated to the user.

4. An image processing apparatus according to claim 2, wherein each time that an object representing a reference is detected, said image processor updates said area parameter so as to set up an area surrounding the reference object, transforms said area parameter designated on the second coordinate system to that on the first coordinate system in accordance with the position of said reference object, and decides whether an object other than said reference object in said area under said image on the first coordinate system, and then the object other than said reference object is indicated to the user.

5. An image processing apparatus according to claim 2, wherein said image processor transforms an image on the first coordinate system to an image on a third coordinate system in which an object'"'"'s height in the transformed image is proportional to an actual height of the object irrespective of the distance from the camera, said designation unit receives setting information as a value on any of the first, second and third coordinate systems, and said image processor decides whether an object is in said area under the second coordinate system and then decides whether the object satisfies said size parameters under the transformed image which has been transformed into the third coordinate system.

6. An image processing apparatus according to claim 5, wherein said setting information has coordinate transformation parameters for transforming the first coordinate system to the second or third coordinate system, and said image processor performs mutual transformation between the first and third coordinate systems through central projection about the principal point using modified transformation parameters with greater depression angle than said transformation parameter.

19. An image processing apparatus according to claim 2, wherein said image processor, when setting information has not been received, calculates differences between a picked-up input image and a background image to obtain a difference image, compares values of respective pixels in a difference image with a predetermined threshold to perform a transformation such that pixels having values larger than the predetermined threshold are transformed to “

255” and

pixels having values less than the predetermined threshold are transformed to “

0”

, to obtain a binarized image, labeling lumps each including “

255”

pixels in the binarized image that are allotted with different numbers to ensure that they can be processed discriminately, updates the background image using the picked-up image, preparing a new background image so that the background image can be kept to be in a latest condition, transforms position, height and width of each of the labeled lumps from those on the first coordinate system to those on the second system based on coordinate transformation parameters, performs an area decision process by referring to the area parameters and when a labeled lump having a height and width which corresponds to the size parameter is present in an area designated by the area parameters, determines the lump as being an intrusion object.

20. An image processing apparatus according to claim 3, wherein when the user defines an area including an object to be detected having a predetermined radius, the image processor provides the picked-up image an overlay display of an elliptical area centered on a position of the object to be detected, sets a label number of the object to be detected and the predetermined radius of the area as area parameters, and based on a size of the object to be detected at the time the user defines the area, sets a size parameter for the object to be detected.

7. An image processing method for processing images picked up by a camera to detect objects in the images, said image processing method comprising the steps of:

processing image data stored in a memory and detecting an object in the processed image data;

displaying the processed image data including the detected object on a display; and

receiving setting information including coordinate transformation parameters and area parameters input by a user as a polygon or a circle defining an area in which a detected object is to be indicated to a user,

wherein said processing step comprises;

transforming an image on a first coordinate system representing a coordinate system during pickup to an image on a second coordinate system set up on a plane representing a reference based on the coordinate transformation parameters,

wherein said displaying displays the transformed image,

wherein said receiving receives said polygon or the circle input by the user on the transformed image being displayed in the screen,

transforming the picked-up image or/and the area parameters by using the coordinate transformation parameters and then detecting an object inside the area defined by the area parameters for indicating to the user, and

wherein said second coordinate system is an approximation of an orthonormal coordinate system having two axes on a ground surface plane so that said image on the second coordinate system analogizes to a map of a monitoring target area, said coordinate transformation parameters includes independent parameters corresponding to a depression angle, an installation height and a view field angle of said camera, and said image processor performs mutual transformation between the first coordinate system and the second coordinate system through central projection about the center at a principal point of the camera.

8. An image processing method according to claim 7, wherein said setting information has size parameters for defining a range of actual height of the object in normal direction to be detected.

9. An image processing method according to claim 8, wherein each time that an object representing a reference is detected, said area parameter is updated to set up an area surrounding the reference object and said image processor decided whether an object other than said reference object is in said area under the second coordinate system, and then the object other than said reference object is indicated to the user.

10. An image processing method according to claim 8, wherein each time that an object representing a reference is detected, said image processor updates said area parameter so as to set up an area surrounding the reference object, transforms said area parameter designated on the second coordinate system to that on the first coordinate system in accordance with the position of said reference object, and decides whether an object other than said reference object in said area under said image on the first coordinate system, and then the object other than said reference object is indicated to the user.

11. An image processing method according to claim 8, wherein said image processor transforms an image on the first coordinate system to an image on a third coordinate system in which an object'"'"'s height in the transformed image is proportional to an actual height of the object irrespective of the distance from the camera, said designation unit receives setting information as a value on any of the first, second and third coordinate systems, and said image processor decides whether an object is in said area under the second coordinate system and then decides whether the object satisfies said size parameters under the transformed image which has been transformed into third coordinate system.

12. An image processing method according to claim 11, wherein said setting information has coordinate transformation parameters for transforming the first coordinate system to the second or third coordinate system, and said image processor performs mutual transformation between the first and third coordination systems through central projection about the principal point using modified transformation parameters with greater depression angle than said transformation parameter.

13. A computer program, stored on a non-transitory computer readable storage medium, for processing images picked up by a camera to detect objects in the images, said computer program, when executed, causes an image processing apparatus to perform the steps of:

processing image data stored in a memory and detecting an object in the processed image data;

displaying the processed image data including the detected object on a display; and

receiving setting information including coordinate transformation parameters and area parameters input by a user as a polygon or a circle defining an area in which a detected object is to be indicated to a user,

wherein said processing step comprises;

transforming an image on a first coordinate system representing a coordinate system during pickup to an image on a second coordinate system set up on a plane representing a reference based on the coordinate transformation parameters,

wherein said displaying displays the transformed image,

wherein said receiving receives said polygon or the circle input by the user on the transformed image being displayed,

transforming the picked-up image or/and the area parameters by using the coordinate transformation parameters and then detects an object inside the area defined by the area parameters for indicating to the user, and

wherein said second coordinate system is an approximation of an orthonormal coordinate system having two axes on a ground surface plane so that said image on the second coordinate system analogizes to a map of a monitoring target area, said coordinate transformation parameters includes three independent parameters corresponding to a depression angle, an installation height and a view field angle of said camera and said image processor performs mutual transformation between the first coordinate system and the second coordinate system through central projection about the center at a principal point of the camera.

14. A computer program according to claim 13, wherein said setting image has size parameters for defining a range of actual height of the object, in normal direction to be detected.

15. A computer program according to claim 14, wherein each time that an object representing a reference is detected, said area parameter is updated to set up an area surrounding the reference object and said image processor decides whether an object other than said reference object is in said area under the second coordinate system, and then the object other than said reference object is indicated to the user.

16. A computer program according to claim 14, wherein each time that an object representing a reference is detected, said image processor updates said area parameter so as to set up an area surrounding the reference object, transforms said area parameter designated on the, second coordinate system to that on the first coordinate system in accordance with the position of said reference object, and decides whether an object other than said reference object in said area under said image on the first coordinate system, and then the object other than said reference object is indicated to the user.

17. A computer program according to claim 14, wherein said image processor transforms an image on the first coordinate system to an image on a third coordinate system in which an object'"'"'s height in the transformed image is proportional to an actual height of the object irrespective of the distance from the camera, said designation unit receives setting information as a value on any of the first, second and third coordinate systems, and said image processor decides whether an object is in said area under the second coordinate system and then decides whether the object satisfies said size parameters under the transformed image which has been transformed into the third coordinate system.

18. A computer program according to claim 17, wherein said setting information has coordinate transformation parameter for transforming the first coordinate system to the second or third coordinate system, and said image processor performs mutual transformation between the first and third coordinate systems through central projection about the principal point using modified transformation parameters with greater depression angle than said transformation parameter.