Matching objects in images
There is provided a method of matching features depicted in images, comprising: detecting a first object depicted in a first image, projecting a first epipolar line, from the first object of the first image, to a second image, selecting second objects along the first epipolar line of the second image, projecting second epipolar lines, from the second objects of the second image, to a third image, projecting a third epipolar line from the first object of the first image to the third image, identifying on the third image, a third object along an intersection of the third epipolar line and a certain second epipolar line of the second epipolar lines, and generating an indication of the first object depicted in first image and the third object depicted in the third image as matches of a same physical object.
1. A computer implemented method of matching features depicted in a plurality of images, comprising:
detecting a first object depicted in a first image, wherein said first image depicting a plurality of objects having matching visual features with said first object;
projecting a first epipolar line, from the first object of the first image, to a second image, wherein said second image depicting at least some of said plurality of objects having matching visual features with said first object;
selecting a plurality of second objects along the first epipolar line of the second image;
projecting a plurality of second epipolar lines, from the plurality of second objects of the second image, to a third image, wherein said third image depicting at least some of said plurality of objects having matching visual features with said first object;
projecting a third epipolar line from the first object of the first image to the third image;
identifying on the third image, a third object along an intersection of the third epipolar line and a certain second epipolar line of the plurality of second epipolar lines; and
generating an indication of the first object depicted in the first image and the third object depicted in the third image as being a same physical object out of the plurality of objects having matching visual features with said first object.
2. The computer implemented method of claim 1 , in response to identifying a plurality of third objects on the third image, further comprising:
obtaining a fourth image of the plurality of images,
projecting a first epipolar line, from the first object of the first image, to the fourth image;
selecting a plurality of second objects along the first epipolar line of the fourth image;
projecting a plurality of second epipolar lines, from the plurality of second objects of the fourth image, to the third image;
wherein the third object is identified on the third image, and the indication is generated for the first, second, third, and fourth images.
3. The computer implemented method of claim 2 , in response to identifying a plurality of third objects on the third image, further comprising iterating the obtaining, the projecting the first epipolar line, the selecting, and the projecting the plurality of second epipolar lines, using sequentially obtained new images until a single third object is identified.
4. The computer implemented method of claim 1 , wherein the first image overlaps the second image and the third image, the second image overlaps the first image and the third image, and the third image overlaps the first image and the second image.
5. The computer implemented method of claim 1 , further comprising mapping the certain second epipolar line of the third image, to a corresponding fourth object of the second image,
wherein the fourth object depicted in the second image is the same as the first object depicted in the first image and the third object depicted in the third image.
6. The computer implemented method of claim 5 , wherein generating the indication further comprises generating the indication of the fourth object depicted in the second image as the same physical object as the first object depicted in the first image and the third object depicted in the third image.
7. The computer implemented method of claim 5 , wherein the fourth object lies along the first epipolar line depicted in the second image.
8. The computer implemented method of claim 5 , wherein the fourth object is one of the plurality of second objects along the first epipolar line of the second image.
9. The computer implemented method of claim 1 , further comprising feeding the indication of the first object depicted in first image and the third object depicted in the third image to an image processing process that processes images based on matching features identified in a plurality of images.
10. The computer implemented method of claim 9 , wherein the image processing process is selected from a group comprising: homography, stereoscopic vision, epipolar related search, 3D reconstruction, and orthophoto creation.
11. The computer implemented method of claim 1 , wherein the plurality of objects that have matching visual features are positioned in at least one of: in near proximity to each other, and arranged in a repeating pattern.
12. The computer implemented method of claim 1 , further comprising accessing parameters of at least one imaging sensor that captured the plurality of images, wherein the epipolar lines are computed according to the parameters, wherein the parameters include at least one of: position of the at least one imaging sensor, orientation of the at least one imaging sensor, and intrinsic parameters of the at least one imaging sensor.
13. The computer implemented method of claim 12 , wherein the parameters are at least one of computed and refined using a bundle adjustment process.
14. The computer implemented method of claim 1 , wherein detecting the first object comprises detecting a plurality of first objects depicted in the first image, wherein the projecting the first epipolar line, the selecting the plurality of second objects, the projecting the plurality of second epipolar lines, the projecting the third epipolar line, the identifying on the third image, and the generating the indication, are iterated for each one of the plurality of first objects.
15. A system for matching features depicted in a plurality of images, comprising:
at least one processor executing a code for:
detecting a first object depicted in a first image, wherein said first image depicting a plurality of objects having matching visual features with said first object;
projecting a first epipolar line, from the first object of the first image, to a second image, wherein said second image depicting at least some of said plurality of objects having matching visual features with said first object;
selecting a plurality of second objects along the first epipolar line of the second image;
projecting a plurality of second epipolar lines, from the plurality of second objects of the second image, to a third image, wherein said third image depicting at least some of said plurality of objects having matching visual features with said first object;
projecting a third epipolar line from the first object of the first image to the third image;
identifying on the third image, a third object along an intersection of the third epipolar line and a certain second epipolar line of the plurality of second epipolar lines; and
generating an indication of the first object depicted in the first image and the third object depicted in the third image as being a same physical object out of the plurality of objects having matching visual features with said first object.
16. A non-transitory medium storing program instructions for matching features depicted in a plurality of images, which, when executed by at least one processor, cause the at least one processor to:
detect a first object depicted in a first image, wherein said first image depicting a plurality of objects having matching visual features with said first object;
project a first epipolar line, from the first object of the first image, to a second image, wherein said second image depicting at least some of said plurality of objects having matching visual features with said first object;
select a plurality of second objects along the first epipolar line of the second image;
project a plurality of second epipolar lines, from the plurality of second objects of the second image, to a third image, wherein said third image depicting at least some of said plurality of objects having matching visual features with said first object;
project a third epipolar line from the first object of the first image to the third image;
identify on the third image, a third object along an intersection of the third epipolar line and a certain second epipolar line of the plurality of second epipolar lines; and
generate an indication of the first object depicted in the first image and the third object depicted in the third image as being a same physical object out of the plurality of objects having matching visual features with said first object.