Detecting a border region in an image
View Patent ↗Systems and methods for detecting a border region in an image. A blank border in a video picture is determined by summing luminance or other pixel measures in a direction parallel to the border and looking for the maximum gradient of those summed measures in a direction perpendicular to the border. Sensitivity can be enhanced by increasing (relative to other pixels) the gain of pixels around the present pixel value of the border. The location of the maximum gradient may be weighted by other measures before a decision on border location is taken.
1. A method for detecting a linear boundary between a border region and a picture region within an image, the method comprising:
measuring a pixel-value spatial gradient in a direction perpendicular to an edge of the image;
identifying a position having the maximum measured pixel-value spatial gradient, wherein the linear boundary is the identified position; and
modifying a spatial-gradient parameter using a distance parameter based upon the distance from the edge of the image, and determining the maximum measured pixel-value spatial gradient using the modified spatial gradient parameter,
wherein the distance parameter is a cumulative sum of high-value pixels.
2. A method according to claim 1 , further comprising combining pixel-values in a direction of the linear boundary, and determining a spatial-gradient parameter of the combined pixel values.
3. A method according to claim 1 , further comprising modifying pixel-values using a nonlinear function which has a maximum slope around a pixel-value corresponding to the black level of the image and a rapidly-decreasing slope for pixel-values increasingly different from that value, and determining the pixel-value spatial gradient using the modified pixel-values.
4. A method according to claim 1 , further comprising combining data from two or more images in a sequence of images, and detecting a border position for an image in the sequence using the combined data.
5. A method according to claim 1 , further comprising using one of a pixel luminance value, a pixel primary colour separation value, and a pixel colour difference value to measure the pixel-value spatial gradient.
6. A method according to claim 1 , further comprising detecting the absence of borders in an image by
locating a horizontal top boundary, a horizontal bottom boundary, a vertical left boundary, and a vertical right boundary; and
determining the absence of borders when each of the horizontal top boundary, the horizontal bottom boundary, the vertical left boundary, and the vertical right boundary coincide with the edges of the image.
7. An apparatus for detecting a linear boundary between a border region and a picture region within an image, comprising:
a spatial gradient measurement module configured to measure a pixel-value spatial gradient in a direction perpendicular to the said boundary; and
a maximum detector module configured to receive the spatial gradient measure and to provide a boundary location;
wherein the maximum detector module also receives a cumulative sum of high-value pixels and the maximum detector module is configured to identify a maximum pixel-value spatial gradient using a weighted sum of the spatial gradient measure and the cumulative sum.
8. An apparatus according to claim 7 , further comprising an averager module configured to combine pixel-values in the direction of the said linear boundary prior to the determination of the spatial gradient.
9. An apparatus according to claim 7 , wherein data from two or more images in a sequence of images is combined and a border position for an image in the sequence is detected using the combined data.
10. A non-transitory computer readable medium encoded with a plurality of computer executable instructions for detecting a linear boundary between a border region and a picture region within an image, the plurality of instructions including instructions for:
measuring a plurality of pixel-value spatial-gradient parameters in a direction perpendicular to the boundary;
detecting a maximum pixel-value spatial-gradient parameter from the measured pixel-value spatial-gradient parameters; and
identifying the boundary as the position of the detected maximum pixel-value spatial gradient;
further comprising modifying a spatial-gradient parameter using a distance Parameter based upon the distance from the edge of the image, and
determining the maximum measured pixel-value spatial gradient using the modified spatial gradient parameter,
wherein the distance parameter is a cumulative sum of high-value pixels.