초록 ▼

A system and method for adjusting exposure in a mosaiced or stitched image. A stitched composite image is typically represented by a set of images and a set of associated transformations. Each transformation corresponds to one image in the input image sequence and represents the mapping between imag

A system and method for adjusting exposure in a mosaiced or stitched image. A stitched composite image is typically represented by a set of images and a set of associated transformations. Each transformation corresponds to one image in the input image sequence and represents the mapping between image pixels in each image and a three-dimensional coordinate system. Every triplet of images in the mosaiced image, having a first, center and third image, is input into the system and method according to the present invention. Before exposure adjustment can be performed, the regions of overlap between the input images in the mosaiced image are calculated. Once the areas of overlap associated with the first and third images are found, the areas of overlap associated with these images are warped to the coordinate frame of the center image. This allows the corresponding pixel locations in the first and center images, and the corresponding pixel location in the third and center images to be determined for the overlapping areas. Additionally, the luminance of each pixel in these areas is identified. The exposure correction factors for each overlapping area are then calculated. Once the exposure correction factors for each overlapping area are calculated, the luminance of all pixels in the center image are adjusted using these factors to yield a mosaiced image that appears to be consistent in exposure. This process is repeated until all sets of triplets have been considered.

대표청구항 ▼

1. A computer-implemented process for improving the uniformity in exposure in a composite image formed from at least a triplet of overlapping images, said process comprising using a computer to perform the following process actions, for each triplet of said images:locating the areas of overlap betwe

1. A computer-implemented process for improving the uniformity in exposure in a composite image formed from at least a triplet of overlapping images, said process comprising using a computer to perform the following process actions, for each triplet of said images:locating the areas of overlap between a first, center and third image;warping the areas of overlap associated with the first and third images to the coordinate frame of the center image;identifying the corresponding pixels in overlapping image areas and their luminance values;computing at least one exposure adjustment parameter for each overlapping area based on the differences between the luminances of each correspondingly located pixel in the overlap area; andadjusting the luminance for every pixel in the center image based on the at least one exposure adjustment parameter associated with each overlap area. 2. The process of claim 1, wherein the process action of computing at least one exposure adjustment parameter for each overlapping area, comprises the actions of:for each pixel location associated with the overlap area between the first and center image, averaging the luminance values of correspondingly located pixels and setting the average luminance value equal to the product of a first gain factor multiplied by the luminance of the center image pixel plus a first bias factor, thereby creating a first series of equations;computing a value for said first gain factor and said first bias factor from said first series of equations;for each pixel location associated with the overlap area between the third and center image, averaging the luminance values of correspondingly located pixels and setting the average luminance value equal to the product of a second gain factor multiplied by the luminance of the center image pixel plus a second bias factor, thereby creating a second series of equations; andcomputing a value for said second gain factor and said second bias factor from said second series of equations. 3. The process of claim 2, wherein the process action of computing a value for said first gain factor and said first bias factor from said first series of equations, comprises the action of employing a least squares approach to compute said first gain and bias values. 4. The process of claim 2, wherein the process action of computing a value for said second gain factor and said second bias factor from said first series of equations, comprises the action of employing a least squares approach to compute said second gain and bias values. 5. The process of claim 1, wherein the process action of adjusting the luminance for every pixel in the center image, comprises the actions of:for each column of pixels in the center image if the triplet of images are lined up horizontally, and for each row of pixels in the center image if the triplet of images are lined up vertically,computing separate gain and bias factors by linearly ramping from the first gain factor to the second gain factor and from the first bias factor to the second bias factor,multiplying the luminance of each pixel of said column or row by the gain factor associated with that column or row, and adding the resulting product to the bias factor associated with the column or row, to compute an adjusted luminance for the pixel under consideration,assigning the adjusted luminance computed for each pixel to that pixel. 6. The process of claim 1, further comprising the process action of blending the transitions between overlapping and non-overlapping input image regions. 7. The process of claim 6, wherein the process action of blending the transitions between overlapping and non-overlapping input image regions further comprises the process action of weighting proportionally to its distance to the edge of the overlapping region the luminance value of each pixel. 8. The process of claim 1, wherein more than one triplet of images from a stitched image is processed to correct the exposure of an entire stitched imag e and wherein a third input image of a given triplet is processed as the first input image of another consecutive triplet. 9. A system for generating a composite image, formed from at least a triplet of overlapping images, that exhibits an improved uniformity in exposure, the system comprising:a general purpose computing device;a computer program comprising program modules executable by the computing device, wherein the computing device is directed by the program modules of the computer program to,for each triplet of images in the composite image,locate the areas of overlap between a first, center and third image;warp the areas of overlap associated with the first and third images to the coordinate frame of the center image;compute a gain factor and a bias factor for each overlapping area based on the differences between the luminances of each correspondingly located pixel in the overlap area; andadjust the luminance for every pixel in the center image based on the gain and bias factors associated with each overlap area. 10. The system of claim 9, wherein the program module for computing the gain and bias factors for each overlapping area, comprises sub-modules for:for each pixel location associated with the overlap area between the first and center image, averaging the luminance values of correspondingly located pixels associated with the pixel location under consideration and setting the average luminance value equal to the product of a first gain factor multiplied by the luminance of the center image pixel associated with the pixel location plus a first bias factor, thereby creating a first series of equations;computing a value for said first gain factor and said first bias factor from said first series of equations using a least squares approach;for each pixel location associated with the overlap area between the third and center image, averaging the luminance values of correspondingly located pixels associated with the pixel location under consideration and setting the average luminance value equal to the product of a second gain factor multiplied by the luminance of the center image pixel plus a second bias factor, thereby creating a second series of equations; andcomputing a value for said second gain factor and said second bias factor from said second series of equations using a least squares approach. 11. The system of claim 9, wherein the program module for adjusting the luminance for every pixel in the center image, comprises sub-modules for:for each column of pixels in the center image if the triplet of images are lined up horizontally,computing separate gain and bias factors by linearly ramping from the first gain factor to the second gain factor and from the first bias factor to the second bias factor,multiplying the luminance of each pixel of said column by the gain factor associated with that column, and adding the resulting product to the bias factor associated with the column, to compute an adjusted luminance for the pixel under consideration, and assigning the adjusted luminance computed for each pixel to that pixel. 12. The system of claim 9, wherein the program module for adjusting the luminance for every pixel in the center image, comprises sub-modules for:for each row of pixels in the center image if the triplet of images are lined up vertically,computing separate gain and bias factors by linearly ramping from the first gain factor to the second gain factor and from the first bias factor to the second bias factor,multiplying the luminance of each pixel of said row by the gain factor associated with that row, and adding the resulting product to the bias factor associated with the row, to compute an adjusted luminance for the pixel under consideration, andassigning the adjusted luminance computed for each pixel to that pixel. 13. The system of claim 9, further comprising a program module to feather the transitions between overlapping and non-overlapping input image regions. 14. A computer-readable medium h aving computer-executable instructions for improving the uniformity in exposure in a composite image formed from at least a triplet of overlapping images, said computer-executable instructions comprising:for each triplet of images in the composite image,locating the areas of overlap between a first, center and third image;warping the areas of overlap associated with the first and third images to the coordinate frame of the center image;computing at least one exposure adjustment parameter for each overlapping area based on the differences between the luminances of each correspondingly located pixel in the overlap area; andadjusting the luminance for every pixel in the center image based on the at least one exposure adjustment parameter associated with each overlap area. 15. The computer-readable medium of claim 14 wherein the instructions for computing at least one exposure adjustment parameter for each overlapping area, comprises sub-modules to:for each pixel location associated with the overlap area between the first and center image, average the luminance values of correspondingly located pixels and set the average luminance value equal to the product of a first gain factor multiplied by the luminance of the center image pixel plus a first bias factor, to thereby create a first series of equations;compute a value for said first gain factor and said first bias factor from said first series of equations;for each pixel location associated with the overlap area between the third and center image, average the luminance values of correspondingly located pixels and set the average luminance value equal to the product of a second gain factor multiplied by the luminance of the center image pixel plus a second bias factor, to thereby create a second series of equations; andcompute a value for said second gain factor and said second bias factor from said second series of equations. 16. The computer-readable medium of claim 14 wherein the instructions for computing a value for said first gain factor and said first bias factor from said first series of equations, comprises a sub-module for employing a least squares approach to compute said first gain and bias values. 17. The computer-readable medium of claim 14, wherein the instructions for adjusting the luminance for every pixel in the center image, comprises sub-modules to:for each row of pixels in the center image of a triplet of images that are composited horizontally,compute individual gain and bias factors for each column of the center image by linearly ramping from the first gain factor to the second gain factor and from the first bias factor to the second bias factor,respectively multiply the luminance of each pixel of the row under consideration by the gain factor associated with the column to which the pixel belongs, and add the resulting product to the bias factor associated with the column to which the pixel belongs, to compute an adjusted luminance for each pixel of the row under consideration; andassign the adjusted luminance computer for each pixel to that pixel. 18. The computer-readable medium of claim 14, wherein the instructions for adjusting the luminance for every pixel in the center image, comprises sub-modules to:for each column of pixels in the center image of a triplet of images that are composited vertically,compute individual gain and bias factors for each row of the center image by linearly ramping from the first gain factor to the second gain factor and from the first bias factor to the second bias factor,respectively multiply the luminance of each pixel of the column under consideration by the gain factor associated with the row to which the pixel belongs, and add the resulting product to the bias factor associated with the row to which the pixel belongs, to compute an adjusted luminance for each pixel of the column under consideration; andassign the adjusted luminance computer for each pixel to that pixel. 19. A system for generating a panoramic video of a surrounding scene comprising:a camera rig comprising multiple video cameras disposed so as to view different portions of the scene, but wherein the field of view of each camera overlaps that of each adjacent camera;a computer program comprising program modules executable by the computing device, wherein the computing device is directed by the program modules of the computer program to,input multiple videos of the scene, wherein each was captured using a separate one of the cameras of the camera rig,stitching together individual frames of the inputted videos which were captured at approximately the same moment in time to form each panoramic frame of the panoramic video;for each triplet of images used to form the panoramic frame,locate the areas of overlap between a first, center and third image;warp the areas of overlap associated with the first and third images to the coordinate frame of the center image;identify the corresponding pixels in overlapping image areas and their luminance values;compute at least one exposure adjustment parameter for each overlapping area based on the differences between the luminances of each correspondingly located pixel in the overlap area; andadjust the luminance for every pixel in the center image based on the at least one exposure adjustment parameter associated with each overlap area.