Image Classification for Adjustment
Image classification techniques are described for adjustment of an image. In one or more implementations, an image is classified by one or more computing device based on suitability of the image for adjustment to correct perspective distortion of the image. Responsive to a classification of the image as not suitable for the adjustment, suitability of the image is detected for processing by a different image adjustment technique by the one or more computing devices.
1 . A method implemented by one or more computing devices, the method comprising:
classifying an image based on suitability of the image for adjustment by an image adjustment technique to correct perspective distortion of the image; and
responsive to the classifying of the image as not suitable for the adjustment, detecting suitability of the image for processing by a different image adjustment technique.
2 . A method as described in claim 1 , further comprising responsive to the classifying of the image as suitable for the adjustment, performing one or more adjustment operations of the image adjustment technique on the image.
3 . A method as described in claim 1 , wherein the image adjustment technique includes upright adjustment.
4 . A method as described in claim 1 , wherein the classifying of the image as suitable for adjustment by the image adjustment technique further comprises analyzing edge pixels and line pixels of the image.
5 . A method as described in claim 4 , wherein the analyzing is based on a ratio between a number of the edge pixels and a number of the line pixels.
6 . A method as described in claim 1 , wherein the classifying of the image as suitable for adjustment by the image adjustment technique further comprises analyzing line pixels and vanishing line pixels of the image.
7 . A method as described in claim 6 , wherein the analyzing is based on a ratio between a number of the line pixels and a number of the vanishing line pixels.
8 . A method as described in claim 1 , wherein the classifying of the image as suitable for adjustment by the image adjustment technique further comprises analyzing vertical and horizontal vanishing lines in the image.
9 . A method as described in claim 8 , wherein the analyzing is based on an average of line segment lengths of the vertical and the horizontal vanishing lines in the image.
10 . A method as described in claim 1 , wherein the detecting of the suitability for the different image adjustment technique includes detecting whether a horizon is identifiable from the image.
11 . A method as described in claim 10 , further comprising:
responsive to the detecting that the horizon is identifiable from the image, performing one or more horizon adjustment operations on the image; and
responsive to the detecting that the horizon is not identifiable from the image, skipping performance of the one or more horizon adjustment operations on the image.
12 . One or more computer-readable storage media comprising instructions stored thereon that, responsive to execution by a computing device, causes the computing device to perform operations comprising:
classifying an image based on suitability of the image for upright adjustment, the classifying comprising:
analyzing edge pixels and line pixels in the image;
analyzing line pixels and vanishing line pixels in the image; or
analyzing vertical and horizontal vanishing lines in the image; and
responsive to the classifying of the image as suitable for upright adjustment, performing one or more upright adjustment operations on the image that correct perspective distortion of the image.
13 . One or more computer-readable storage media as described in claim 12 , wherein the analyzing is based on a ratio between a number of the edge pixels and a number of the line pixels.
14 . One or more computer-readable storage media as described in claim 12 , wherein the analyzing is based on a ratio between a number of the line pixels and a number of the vanishing line pixels.
15 . One or more computer-readable storage media as described in claim 12 , wherein the analyzing is based on an average of line segment lengths of the vertical and the horizontal vanishing lines in the image.
16 . One or more computer-readable storage media as described in claim 12 , further comprising responsive to the classifying of the image as not suitable for the upright adjustment, detecting whether a horizon is identifiable from the image.
17 . A system comprising:
at least one module implemented at least partially in hardware, the at least one module configured to classify an image based on suitability of the image for upright adjustment; and
one or more modules implemented at least partially in hardware, the one or more modules configured to perform one or more upright adjustment operations on the image that correct perspective distortion of the image responsive to the classification of the image as suitable for upright adjustment.
18 . A system as described in claim 17 , wherein the at least one module is configured to perform the classification of the image by:
analyzing edge pixels and line pixels of the image;
analyzing line pixels and vanishing line pixels of the image; or
analyzing vertical and horizontal vanishing lines in the image.
19 . A system as described in claim 18 , wherein:
the analyzing is based on a ratio between a number of the edge pixels and a number of the line pixels;
the analyzing is based on a ratio between a number of the line pixels and a number of the vanishing line pixels; or
the analyzing is based on an average of line segment lengths of the vertical and the horizontal vanishing lines in the image.
20 . A system as described in claim 17 , further comprising a module implemented at least partially in hardware and configured to detect whether a horizon is identifiable from the image responsive to the classification of the image as not suitable for the upright adjustment and responsive to the detection that the horizon is identifiable, perform one or more horizon adjustment techniques to the image.