Image alignment using corner and line features

A computer system configured to align content from images that are generated by different cameras, where the alignment is performed by concurrently identifying corresponding corner features and line features within the images and by aligning the images based on a combination of the identified corner features and line features, thereby resulting in relaxed constraints during said alignment as compared to constraints that are required when aligning the images using only the corner features, said computer system comprising:

• one or more processors;
• and
• one or more computer-readable hardware storage devices that store instructions that are executable by the one or more processors to cause the computer system to:
• acquire a first image generated by a first camera;
• acquire a second image generated by a second camera that is different than the first camera;
• detect, within both the first image and the second image, a first set of pixels that correspond to corner features;
• at least partially in parallel with detecting the first set of pixels, detect, within both the first image and the second image, a second set of pixels that correspond to line features;
• and
• align image content from the first image with corresponding image content from the second image, wherein said aligning is performed using at least some of the pixels in the first set of pixels, which correspond to the corner features, and using at least some of the pixels in the second set of pixels, which correspond to the line features.

A method for aligning content from images that are generated by different cameras, wherein the alignment is performed by concurrently identifying unambiguous features that are uniquely identifiable within the images and identifying potentially ambiguous features that are susceptible to aperture ambiguity, and wherein the alignment is based on a combination of the unambiguous features and the potentially ambiguous features, thereby resulting in relaxed constraints during said alignment as compared to constraints that are required when aligning the images using only the unambiguous features, said method comprising:

• acquiring a first image generated by a first camera;
• acquiring a second image generated by a second camera that is detached from the first camera;
• performing a first detection operation that includes:
• detecting, within the first image, a first set of pixels that correspond to unambiguous features that are uniquely identifiable within the first image;
• detecting, within the second image, a second set of pixels that correspond to the unambiguous features that are included in the first image;
• performing a second detection operation that includes:
• detecting, within the first image, a third set of pixels that correspond to potentially ambiguous features that are susceptible to aperture ambiguity;
• detecting, within the second image, a fourth set of pixels that correspond to the potentially ambiguous features that are included in the first image;
• performing an alignment process in which image content from the first image is aligned with corresponding image content from the second image, wherein said alignment process includes:
• reprojecting at least some pixels in the second set of pixels to positions corresponding to positions of at least some pixels in the first set of pixels;
• and
• reprojecting at least some pixels in the fourth set of pixels to positions corresponding to positions of at least some pixels in the third set of pixels.