Image processing apparatus and image processing method

Patent metadata

Field	Value
Publication number	US-9514382-B2
Application number	US-201414341320-A
Country	US
Kind code	B2
Filing date	Jul 25, 2014
Priority date	Aug 28, 2013
Publication date	Dec 6, 2016
Grant date	Dec 6, 2016

Abstract

A binarization processing unit generates a binary image by comparing the visual feature of each pixel of an input image with a predetermined binarization threshold and binarizing the value of each pixel of the input image based on the comparison result. The binarization processing unit performs this process while changing the binarization threshold, thereby generating a plurality of binary images. A contour detecting unit calculates the area of a closed region having an overlap between the plurality of binary images, and determines the binarization threshold to be used for detecting a boundary of the corresponding closed region as a contour of an object, based on the area change rate of the corresponding closed region caused by changing the binarization threshold.

First claim

What is claimed is:

1. An image processing apparatus comprising:
a processor configured to perform a procedure including:
generating a plurality of binary images by performing a binary image generation process while changing a predetermined binarization threshold, the binary image generation process being a process of generating a binary image by comparing a visual feature of each pixel of an image representing a plurality of objects with the binarization threshold and binarizing a value of each pixel of the image based on a result of the comparing,
calculating an area of a corresponding closed region, the corresponding closed region being a closed region represented in each of the plurality of binary images and having an overlap between the plurality of binary images,
determining the binarization threshold to be used for detecting a boundary of the corresponding closed region as a contour of an object, based on an area change rate of the corresponding closed region caused by changing the binarization threshold, and
detecting a boundary of a first corresponding closed region as the contour of the object, the first corresponding closed region being represented in a binary image generated by using the determined binarization threshold among the plurality of binary images.

2. The image processing apparatus according to claim 1 , wherein:
the generating a plurality of binary images includes changing the binarization threshold such that the corresponding closed region is gradually enlarged; and
the determining includes, when the area change rate of the corresponding closed region becomes equal to or greater than a predetermined maximum threshold between two binary images that are consecutively generated while changing the binarization threshold, detecting a boundary of the corresponding closed region in a preceding one of the two binary images as the contour of the object.

3. The image processing apparatus according to claim 2 , wherein the procedure further includes, when the area change rate of the corresponding closed region becomes less than a predetermined minimum threshold between the two binary images, the minimum threshold being less than the maximum threshold, detecting a boundary of the corresponding closed region in either one of the two binary images as the contour of the object.

4. The image processing apparatus according to claim 1 , wherein:
the procedure further includes generating a visual feature image, by back-projecting a frequency of a value that is based on the pixel value of each pixel of the image onto the pixel of the image; and
the generating a plurality of binary images includes generating the binary image by comparing a value of each pixel of the visual feature image with the binarization threshold.

5. The image processing apparatus according to claim 4 , wherein:
the procedure further includes performing distance conversion on the visual feature image, and extracting maximal points in an image obtained by the distance conversion as reference points of the respective objects; and
the calculating includes calculating, as the area of the corresponding closed region, areas of closed regions represented in the respective plurality of binary images and including a same one of the reference points.

6. The image processing apparatus according to claim 1 , wherein:
the procedure further includes extracting reference points of the respective objects, based on the visual feature of each pixel in the image; and
the calculating includes calculating, as the area of the corresponding closed region, areas of closed regions represented in the respective plurality of binary images and including a same one of the reference points.

7. The image processing apparatus according to claim 6 , wherein the determining includes:
dividing a surrounding region of the reference point into a plurality of sub-regions,
calculating an area of a partial closed region, the partial closed region being represented in each of the sub-regions and forming a part of the corresponding closed region, and
determining, for each sub-region, the binarization threshold to be used for detecting a boundary of the partial closed region as a part of the contour of the object, based on an area change rate of the partial closed region caused by changing the binarization threshold.

8. An image processing method comprising:
generating a plurality of binary images by performing a binary image generation process while changing a predetermined binarization threshold, the binary image generation process being a process of generating a binary image by comparing a visual feature of each pixel of an image representing a plurality of objects with the binarization threshold and binarizing a value of each pixel of the image based on a result of the comparing;
calculating an area of a corresponding closed region, the corresponding closed region being a closed region represented in each of the plurality of binary images and having an overlap between the plurality of binary images;
determining the binarization threshold to be used for detecting a boundary of the corresponding closed region as a contour of an object, based on an area change rate of the corresponding closed region caused by changing the binarization threshold; and
detecting a boundary of a first corresponding closed region as the contour of the object, the first corresponding closed region being represented in a binary image generated by using the determined binarization threshold among the plurality of binary images.

9. The image processing method according to claim 8 , wherein:
the generating a plurality of binary images includes changing the binarization threshold such that the corresponding closed region is gradually enlarged; and
the determining includes, when the area change rate of the corresponding closed region becomes equal to or greater than a predetermined maximum threshold between two binary images that are consecutively generated while changing the binarization threshold, detecting a boundary of the corresponding closed region in a preceding one of the two binary images as the contour of the object.

10. The image processing method according to claim 9 , further comprising:
when the area change rate of the corresponding closed region becomes less than a predetermined minimum threshold between the two binary images, the minimum threshold being less than the maximum threshold, detecting a boundary of the corresponding closed region in either one of the two binary images as the contour of the object.

11. The image processing method according to claim 8 , further comprising:
generating a visual feature image, by back-projecting a frequency of a value that is based on the pixel value of each pixel of the image onto the pixel of the image;
wherein the generating a plurality of binary images includes generating the binary image by comparing a value of each pixel of the visual feature image with the binarization threshold.

12. The image processing method according to claim 11 , further comprising:
performing distance conversion on the visual feature image, and extracting maximal points in an image obtained by the distance conversion as reference points of the respective objects;
wherein the calculating includes calculating, as the area of the corresponding closed region, areas of closed regions represented in the respective plurality of binary images and including a same one of the reference points.

13. The image processing method according to claim 8 , further comprising:
extracting reference points of the respective objects, based on the visual feature of each pixel in the image;
wherein the calc

Assignees

• Fujitsu Ltd

Inventors

• Ishihara Masaki
• Sugimura Masahiko
• Endo Susumu
• Baba Takayuki
• Uehara Yusuke

Classifications

• G06V10/46 Primary

  Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features (colour feature extraction G06V10/56) · CPC title

• G06T7/11 Primary

  Region-based segmentation · CPC title

• G06V10/26

  Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion · CPC title

• G06T2207/20148

  Physics · mapped topic

• G06T7/0081

  Physics · mapped topic