Back to results
US 8,285,002 B2Granted
Image processing apparatus and method, image sensing apparatus, and program
AssigneeCanon Kabushiki Kaisha
InventorsKatsuhiko Mori, Masakazu Matsugu, Yusuke Mitarai, Yuji Kaneda
Life of this patent
Estimating the term…
Abstract
An eye region is detected, in the detected eye region, the region of an iris outline is extended inward, and a hollow region surrounded by the extended region is detected as a poor hue quality region. An outline of the poor hue quality region is shaped by approximating the outline to one of a circle and an ellipse. The hue of the shaped poor hue quality region is corrected on the basis of color information obtained from a neighboring region inside an iris outline.
Claims12
An image processing method comprising:
- an input step of inputting an image;
- an eye region detection step of detecting an eye region from the input image;
- an iris outline detection step of detecting an iris outline from the eye region;
- an iris color setting step of setting an iris color based on a color distribution of pixels having luminance values of a first predetermined ratio in a neighboring region inside the iris outline;
- a region extension step of executing region extension from a reference pixel inside the iris outline so that colors of more than a second predetermined ratio of adjacent pixels in an extended region fall within a predetermined color range with respect to the set iris color, and setting the extended region as a poor hue quality region in the eye region;
- a poor hue quality region shaping step of shaping the poor hue quality region;
- and
- a poor hue quality region correction step of correcting a hue of the shaped poor hue quality region.
5 more elements
An image processing apparatus comprising:
- input means for inputting an image;
- eye region detection means for detecting an eye region from the input image;
- iris outline detection means for detecting an iris outline from the eye region;
- iris color setting means for setting an iris color based on a color distribution of pixels having luminance values of a first predetermined ratio in a neighboring region inside the iris outline;
- region extension means for extending a region from a reference pixel inside the iris outline so that colors of more than a second predetermined ratio of adjacent pixels in an extended region fall within a predetermined color range with respect to the set iris color, and setting the extended region as a poor hue quality region in the eye region;
- poor hue quality region shaping means for shaping the poor hue quality region;
- and
- poor hue quality region correction means for correcting a hue of the shaped poor hue quality region.
5 more elements
An image processing method comprising:
- an input step of inputting an image;
- an eye region detection step of detecting an eye region from the input image;
- an iris outline detection step of detecting an iris outline from the eye region;
- an iris color setting step of setting an iris color based on a color distribution of pixels having luminance values of a first predetermined ratio in a neighboring region inside the iris outline;
- a region extension step of executing region extension from a reference pixel inside the iris outline so that colors of more than a second predetermined ratio of adjacent pixels in an extended region fall within a predetermined color range with respect to the set iris color, and setting the extended region as a poor hue quality region in the eye region, wherein the colors comprise R, G and B values;
- a poor hue quality region shaping step of shaping the poor hue quality region;
- a poor hue quality region correction step of correcting a hue of the shaped poor hue quality region;
- and
- a non-correction region detection step of detecting a non-correction region from the eye region, wherein the non-correction region is an eyelid region,
- wherein correction processing in the poor hue quality region correction step is executed for a region except the non-correction region detected in the non-correction region detection step.
7 more elements
Dependencies are inferred from the claim text.