Motion detection method and device, program and vehicle surveillance system
First Claim
Patent Images
1. A motion detection method comprising the steps of:
- digitizing a background differential image corresponding to a differential image between an input image sequentially input for each frame and a preset background image so as to generate a digitized image composed of a target region different from the background image and a background region with no difference from the background image for each frame;
determining an overlap region when a value of a common pixel in a current frame is equal to the value of the same pixel in a frame directly preceding the current frame;
determining an increased region when the value of a common pixel in the current frame is greater than the value of the same pixel in the frame directly preceding the current framed;
determining a decreased region when the value of a common pixel in the current frame is less than the value of the same pixel in the frame directly preceding the current frame;
comparing digitized images between frames to extract the overlap region remaining as the target region, the increased region changed from the background region to the target region, and the decreased region changed from the target region to the background region;
when the increased region and the decreased region are adjacent to the overlap region, obtaining an integrated region by grounding the increased region, the decreased region, and the overlap region;
for each of the integrated region, the increased region, and the decreased region, obtaining region parameters composed of the number of pixels contained in the region, a sum of X-coordinates obtained by adding X-coordinate values in two-dimensional coordinates indicating a position of a pixel for all the pixels contained in the region, and a sum of Y-coordinates obtained by adding Y-coordinate values for all the pixels contained in the region; and
detecting a motion of a moving object displayed on the integrated region for each of the integrated regions based on the region parameters.
1 Assignment
0 Petitions
Accused Products
Abstract
Based on an input image from an imaging device and a preset background image, a digitized image is generated for each frame. The digitized images are compared. Then, an increased region, a decreased region, and an overlap region, which are adjacent to each other, are extracted as a single integrated region. Region parameters, which are a region area, a sum of X-coordinates and a sum of Y-coordinates, are obtained for each of the extracted integrated region, the increased region and the decreased region. Based on the region parameters, a motion of the object displayed on the integrated region is obtained for each of the integrated regions.
30 Citations
9 Claims
-
1. A motion detection method comprising the steps of:
-
digitizing a background differential image corresponding to a differential image between an input image sequentially input for each frame and a preset background image so as to generate a digitized image composed of a target region different from the background image and a background region with no difference from the background image for each frame; determining an overlap region when a value of a common pixel in a current frame is equal to the value of the same pixel in a frame directly preceding the current frame;
determining an increased region when the value of a common pixel in the current frame is greater than the value of the same pixel in the frame directly preceding the current framed;determining a decreased region when the value of a common pixel in the current frame is less than the value of the same pixel in the frame directly preceding the current frame; comparing digitized images between frames to extract the overlap region remaining as the target region, the increased region changed from the background region to the target region, and the decreased region changed from the target region to the background region; when the increased region and the decreased region are adjacent to the overlap region, obtaining an integrated region by grounding the increased region, the decreased region, and the overlap region; for each of the integrated region, the increased region, and the decreased region, obtaining region parameters composed of the number of pixels contained in the region, a sum of X-coordinates obtained by adding X-coordinate values in two-dimensional coordinates indicating a position of a pixel for all the pixels contained in the region, and a sum of Y-coordinates obtained by adding Y-coordinate values for all the pixels contained in the region; and detecting a motion of a moving object displayed on the integrated region for each of the integrated regions based on the region parameters.
-
-
2. A motion detection device comprising:
-
digitized image generating means for digitizing a background differential image corresponding to a differential image between an input image sequentially input for each frame and a preset background image so as to generate a digitized image composed of a target region different from the background image and a background region with no difference from the background image for each frame; frame comparison means for determining an overlap region when a value of a common pixel in a current frame is equal to the value of the same pixel in a frame directly preceding the current frame, for determining an increased region when the value of a common pixel in the current frame is greater than the value of the same pixel in the frame directly preceding the current frame, and for determining a decreased region when the value of a common pixel in the current frame is less than the value of the same pixel in the frame directly preceding the current frame; region extracting means for comparing digitized images generated in the digital image generating means between frames to extract the overlap region remaining as the target region, the increased region changed from the background region to the target region, and the decreased region changed from the target region to the background region and, when the increased region and the decreased region are adjacent to the overlap region, for obtaining an integrated region by grouping the increased region, the decreased region, and the overlap region; region parameter calculating means for, for each of the integrated region, the increased region, and the decreased region, obtaining region parameters composed of the number of pixels contained in the region, a sum of X-coordinates obtained by adding X-coordinate values in two-dimensional coordinates indicating a position of a pixel for all the pixels contained in the region, and a sum of Y-coordinates obtained by adding Y-coordinate values for all the pixels contained in the region; and motion detecting means for detecting a motion of a moving object displayed on the integrated region for each of the integrated regions based on the region parameters. - View Dependent Claims (3)
-
-
4. A vehicle surveillance system comprising:
-
digitized image generating means for digitizing a background differential image corresponding to a differential image between an input image sequentially input for each frame and a preset background image so as to generate a digitized image composed of a target region different from the background image and a background region with no difference from the background image for each frame; frame comparison means for determining an overlap region when a value of a common pixel in a current frame is equal to the value of the same pixel in a frame directly preceding the current frame, for determining an increased region when the value of a common pixel in the current frame is greater than the value of the same pixel in the frame directly preceding the current frame, and for determining a decreased region when the value of a common pixel in the current frame is less than the value of the same pixel in the frame directly preceding the current frame; region extracting means for comparing digitized images generated in the digital image generating means between frames to extract the overlap region remaining as the target region, the increased region changed from the background region to the target region, and the decreased region changed from the target region to the background region and, when the increased region and the decreased region are adjacent to the overlap region, for obtaining an integrated region by grouping the increased region, the decreased region, and the overlap region; region parameter calculating means for, for each of the integrated region, the increased region, and the decreased region, obtaining region parameters composed of the number of pixels contained in the region, a sum of X-coordinates obtained by adding X-coordinate values in two-dimensional coordinates indicating a position of a pixel for all the pixels contained in the region, and a sum of Y-coordinates obtained by adding Y-coordinate values for all the pixels contained in the region; motion detecting means for detecting a motion of a moving object displayed on the integrated region for each of the integrated regions based on the region parameters;
imaging means provided in a vehicle compartment so as to supply an image captured by taking the vehicle compartment or the vicinity of the vehicle through a window to the motion detection device as the input image;determination means for determining a presence of any of an intruder and a suspicious individual based on a result of detection in the motion detection device; and
notification means for notifying a result of determination in the determination means. - View Dependent Claims (5, 6, 7, 8)
-
-
9. A computer-readable medium comprising instructions being executed by a computer, the instructions including a computer-implemented method for detecting motion, the instructions for:
-
digitizing a background differential image corresponding to a differential image between an input image sequentially input for each frame and a preset background image so as to generate a digitized image composed of a target region different from the background image and a background region with no difference from the background image for each frame; determining an overlap region when a value of a common pixel in a current frame is equal to the value of the same pixel in the frame directly preceding the current frame; determining an increased region when the value of a common pixel in the current frame is greater than the value of the same pixel in the frame directly preceding the current frame; determining a decreased region when the value of a common pixel in the current frame is less than the value of the same pixel in the frame directly preceding the current frame; comparing digitized images between frames to extract the overlap region remaining as the target region, the increased region changed from the background region to the target region, and the decreased region changed from the target region to the background region; when the increased region and decreased region are adjacent to the overlap region, obtaining an integrated region by grouping the increased region, the decreased region, and the overlap region; for each of the integrated region, the increased region, and the decreased region, obtaining region parameters composed of the number of pixels contained in the region, a sum of X-coordinates obtained by adding X-coordinate values in two-dimensional coordinates indicating a position of a pixel for all the pixels contained in the region, and a sum of Y-coordinates obtained by adding Y-coordinate values for all the pixels contained in the region; and detecting a motion of a moving object displayed on the integrated region for each of the integrated regions based on the region parameters.
-
Specification