초록 ▼

Binary image stitching based on grayscale approximation. In a first example embodiment, a method for automatic binary image stitching of first and second binary images includes several acts. First, the first and second binary images are converted into first and second thumbnail grayscale images. Eac

Binary image stitching based on grayscale approximation. In a first example embodiment, a method for automatic binary image stitching of first and second binary images includes several acts. First, the first and second binary images are converted into first and second thumbnail grayscale images. Each thumbnail grayscale image is a lower resolution than its corresponding binary image by a scaling factor. Next, a thumbnail transform is calculated between the first thumbnail grayscale image and the second thumbnail grayscale image. Then, a full-size transform is calculated as a function of the thumbnail transform and the scaling factor. Next, the stitched binary image is created by projecting the first binary image onto the second binary image using the full-size transform and combining overlapping portions of the first and second binary images.

대표청구항 ▼

1. A method for automatic binary image stitching of first and second binary images, the method comprising the following acts: a) converting the first and second binary images into first and second thumbnail grayscale images, wherein each thumbnail grayscale image is a lower resolution than its corre

1. A method for automatic binary image stitching of first and second binary images, the method comprising the following acts: a) converting the first and second binary images into first and second thumbnail grayscale images, wherein each thumbnail grayscale image is a lower resolution than its corresponding binary image by a scaling factor;b) calculating a thumbnail transform between the first thumbnail grayscale image and the second thumbnail grayscale image;c) calculating a full-size transform as a function of the thumbnail transform and the scaling factor; andd) creating a stitched binary image by projecting the first binary image onto the second binary image using the full-size transform and combining overlapping portions of the first and second binary images. 2. The method as recited in claim 1, wherein the scaling factor comprises a scaling factor k that is the smallest positive integer that satisfies at least one of the following equations: W/k≦Tw, orH/k≦Th;where:W is the width of the binary images;Tw is a predetermined threshold for image width;H is the height of the binary images, andTh is a predetermined threshold for image height. 3. The method as recited in claim 2, wherein the act a) comprises converting the first and second binary images into first and second thumbnail grayscale images according to the following equation: g⁡(x,y)=∑i=0k-1⁢∑j=0k-1⁢w⁡(i,j)×b⁡(x×k+i,y×k+j)where:g(x,y) is the pixel value in y row and x column of the grayscale image;b(i,j) is the pixel value in i row and j column of the binary image;i≦W;j≦H; andw(i,j) is a predetermined weight. 4. The method as recited in claim 3, wherein the predetermined weight w(i,j) satisfies the following equation: ∑i=0k-1⁢∑j=0k-1⁢w⁡(i,j)=1. 5. The method as recited in claim 1, wherein the act b) comprises calculating a thumbnail transform between the first thumbnail grayscale image and the second thumbnail grayscale image using a Harris corner detector. 6. The method as recited in claim 1, wherein the act c) comprises calculating the full-size transform by multiplying a translation distance of the thumbnail transform by the scaling factor while a rotation angle of the thumbnail transform is left unchanged. 7. An image processing apparatus configured to receive first and second binary images and generate a stitched binary image, the image processing apparatus comprising: a grayscale conversion module configured to convert the first and second binary images into first and second thumbnail grayscale images, wherein each thumbnail grayscale image is a lower resolution than its corresponding binary image by a scaling factor;a thumbnail transform module configured to calculate a thumbnail transform between the first thumbnail grayscale image and the second thumbnail grayscale image;a full-size transform module configured to calculate a full-size transform as a function of the thumbnail transform and the scaling factor; anda stitching module configured to create a stitched binary image by projecting the first binary image onto the second binary image using the full-size transform and combining overlapping portions of the first and second binary images. 8. The apparatus as recited in claim 7, wherein the scaling factor comprises a scaling factor k that is the smallest positive integer that satisfies at least one of the following equations: W/k≦Tw, orH/k≦Th;where:W is the width of the binary images;Tw is a predetermined threshold for image width;H is the height of the binary images, andTh is a predetermined threshold for image height. 9. The apparatus as recited in claim 8, wherein the grayscale conversion module is configured to convert the first and second binary images into first and second thumbnail grayscale images according to the following equation: g⁡(x,y)=∑i=0k-1⁢∑j=0k-1⁢w⁡(i,j)×b⁡(x×k+i,y×k+j)where:g(x,y) is the pixel value in y row and x column of the grayscale image;b(i,j) is the pixel value in i row and j column of the binary image;i≦W;j≦H; andw(i,j) is a predetermined weight. 10. The apparatus as recited in claim 9, wherein the predetermined weight w(i,j) satisfies the following equation: ∑i=0k-1⁢∑j=0k-1⁢w⁡(i,j)=1. 11. The apparatus as recited in claim 7, wherein the thumbnail transform module is configured to calculate a thumbnail transform between the first thumbnail grayscale image and the second thumbnail grayscale image using a Harris corner detector. 12. The apparatus as recited in claim 7, wherein the full-size transform module is configured to calculate the full-size transform by multiplying a translation distance of the thumbnail transform by the scaling factor while a rotation angle of the thumbnail transform is left unchanged. 13. The apparatus as recited in claim 7, wherein the apparatus comprises a scanner. 14. One or more non-transitory computer-readable media having computer-readable instructions thereon which, when executed, implement a method for automatic binary image stitching of first and second binary images, the method comprising the acts of: a) converting the first and second binary images into first and second thumbnail grayscale images, wherein each thumbnail grayscale image is a lower resolution than its corresponding binary image by a scaling factor;b) calculating a thumbnail transform between the first thumbnail grayscale image and the second thumbnail grayscale image;c) calculating a full-size transform as a function of the thumbnail transform and the scaling factor;d) creating a stitched binary image by projecting the first binary image onto the second binary image using the full-size transform and combining overlapping portions of the first and second binary images. 15. The one or more non-transitory computer-readable media as recited in claim 14, wherein the scaling factor comprises a scaling factor k that is the smallest positive integer that satisfies at least one of the following equations: W/k≦Tw, orH/k≦Th;where:W is the width of the binary images;Tw is a predetermined threshold for image width;H is the height of the binary images, andTh is a predetermined threshold for image height. 16. The one or more non-transitory computer-readable media as recited in claim 15, wherein the act a) comprises converting the first and second binary images into first and second thumbnail grayscale images according to the following equation: g⁡(x,y)=∑i=0k-1⁢∑j=0k-1⁢w⁡(i,j)×b⁡(x×k+i,y×k+j)where:g(x,y) is the pixel value in y row and x column of the grayscale image;b(i,j) is the pixel value in i row and j column of the binary image;i≦W;j≦H;w(i,j) is a predetermined weight. 17. The one or more non-transitory computer-readable media as recited in claim 16, wherein the predetermined weight w(i,j) satisfies the following equation: ∑i=0k-1⁢∑j=0k-1⁢w⁡(i,j)=1. 18. The one or more non-transitory computer-readable media as recited in claim 14, wherein the act b) comprises calculating a thumbnail transform between the first thumbnail grayscale image and the second thumbnail grayscale image using a Harris corner detector. 19. The one or more non-transitory computer-readable media as recited in claim 14, wherein the act c) comprises calculating the full-size transform by multiplying a translation distance of the thumbnail transform by the scaling factor while a rotation angle of the thumbnail transform is left unchanged. 20. An image processing apparatus comprising: a processor; andthe one or more non-transitory computer-readable media as recited in claim 19, the processor being configured to execute the one or more computer-readable instructions.