Image highlight correction using illumination specific hsv color coordinate
First Claim
Patent Images
1. An image highlight correction method comprising the steps of:
- obtaining the color of a light source to be used in imaging;
illuminating said light source onto an object;
obtaining an RGB reflectance image of said illuminated object using a color imaging device;
plotting the color of the light source as a vector in RGB color space;
constructing a plane, which is perpendicular to the light source vector and passes through the origin of RGB color space;
projecting every pixel of said RGB image along the light source vector onto said perpendicular plane;
constructing an illumination specific HSV color coordinate system by making the origin of the RGB color space be the origin of said illumination specific HSV color coordinate system, projecting the unit base vector, RGB(0, 0,
1), along said light source vector onto said perpendicular plane, making the axis, which connects said projected point and said origin, be the x-axis of said illumination specific HSV color coordinate system, making the axis perpendicular to said x-axis and lying in said perpendicular plane be the y-axis of said illumination specific HSV color coordinate system, and making the axis, which passes through said origin and is perpendicular to said x-axis and y-axis, be the z-axis of said illumination specific HSV color coordinate system;
evaluating the validity of each pixel based on the criterion that the RGB coordinates of a valid pixel are unsaturated, that is, within said color imaging device'"'"'s dynamic range;
converting the coordinates of the projected RGB image points on the x-y plane into polar coordinates, for which the θ
coordinate corresponds to the hue value and the r coordinate corresponds to the saturation value, by setting the positive side of said x-axis to correspond to θ
=0° and
determining the r coordinate as the length of the vector from the origin to a projected point and the θ
coordinate as the counterclockwise angle in degrees with respect to θ
=0°
;
converting the r, θ
coordinates to saturation (S) and hue (H) values by projecting the RGB coordinates of the three primary colors RED, GREEN, and BLUE onto said plane, determining the r, θ
coordinates of the projected points of the three primary colors, determining scaling factors for the respective primary colors and an offset for one primary color such as will provide a correlation of the r, θ
coordinates of the primary colors with the known saturation and hue values for the primary colors, and applying said scaling factors and offset to the r, θ
coordinates of the projected image pixel points to obtain the correct saturation and hue values, respectively, of the projected image pixel points;
determining the non-highlight value coordinate V of each valid pixel; and
estimating the true RGB coordinates of each pixel from said S, H, and V values of each pixel.
0 Assignments
0 Petitions
Accused Products
Abstract
An image highlight correction method comprising the steps of mapping the image data from RGB color coordinates to an illumination specific HSV color coordinate system the value axis of which is in the direction of the light source, estimating highlight free HSV coordinates for each pixel, and restoring a highlight free image in RGB representation. Image highlight correction programs and image acquisition systems that make use of this image highlight correction method are also disclosed.
-
Citations
50 Claims
-
1. An image highlight correction method comprising the steps of:
-
obtaining the color of a light source to be used in imaging;
illuminating said light source onto an object;
obtaining an RGB reflectance image of said illuminated object using a color imaging device;
plotting the color of the light source as a vector in RGB color space;
constructing a plane, which is perpendicular to the light source vector and passes through the origin of RGB color space;
projecting every pixel of said RGB image along the light source vector onto said perpendicular plane;
constructing an illumination specific HSV color coordinate system by making the origin of the RGB color space be the origin of said illumination specific HSV color coordinate system, projecting the unit base vector, RGB(0, 0,
1), along said light source vector onto said perpendicular plane, making the axis, which connects said projected point and said origin, be the x-axis of said illumination specific HSV color coordinate system, making the axis perpendicular to said x-axis and lying in said perpendicular plane be the y-axis of said illumination specific HSV color coordinate system, and making the axis, which passes through said origin and is perpendicular to said x-axis and y-axis, be the z-axis of said illumination specific HSV color coordinate system;
evaluating the validity of each pixel based on the criterion that the RGB coordinates of a valid pixel are unsaturated, that is, within said color imaging device'"'"'s dynamic range;
converting the coordinates of the projected RGB image points on the x-y plane into polar coordinates, for which the θ
coordinate corresponds to the hue value and the r coordinate corresponds to the saturation value, by setting the positive side of said x-axis to correspond to θ
=0° and
determining the r coordinate as the length of the vector from the origin to a projected point and the θ
coordinate as the counterclockwise angle in degrees with respect to θ
=0°
;
converting the r, θ
coordinates to saturation (S) and hue (H) values by projecting the RGB coordinates of the three primary colors RED, GREEN, and BLUE onto said plane, determining the r, θ
coordinates of the projected points of the three primary colors, determining scaling factors for the respective primary colors and an offset for one primary color such as will provide a correlation of the r, θ
coordinates of the primary colors with the known saturation and hue values for the primary colors, and applying said scaling factors and offset to the r, θ
coordinates of the projected image pixel points to obtain the correct saturation and hue values, respectively, of the projected image pixel points;
determining the non-highlight value coordinate V of each valid pixel; and
estimating the true RGB coordinates of each pixel from said S, H, and V values of each pixel. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
-
-
25. An image highlight correction program, comprising
a light source color vector acquisition subroutine, in which the color of a light source is obtained as a vector in RGB color space; -
an object RGB image acquisition subroutine, in which an RGB image of an object illuminated by said light source is obtained;
a pixel validity evaluation subroutine, in which the validity of each pixel is evaluated based on the criterion that the RGB coordinates of a valid pixel are within the color imaging device'"'"'s dynamic range, an HSV color coordinate plane projection subroutine, in which every pixel of said RGB image are projected, along the light source vector, onto the x-y plane of an illumination specific HSV color coordinate system, with which the x-y plane is defined as the plane that is perpendicular to the light source vector and passes through the origin of RGB color space, the origin is defined by said origin of RGB space, the x-axis is defined by the axis, which connects said origin and the point of projection of the unit base vector, RGB(0, 0,
1), along said light source vector onto said x-y plane, the y-axis is defined as the axis perpendicular to said x-axis and lying in said perpendicular plane, and the z-axis is defined as the axis, which passes through said origin and is perpendicular to said x-axis and y-axis;
an r, θ
coordinate conversion subroutine, in which the coordinates of the projected points on the x-y plane are converted into polar coordinates, with which the θ
coordinate corresponds to the hue value and the r coordinate corresponds to the saturation value, by setting the positive side of said x-axis to correspond to θ
=0° and
determining the r coordinate as the length of the vector from the origin to a projected point and the θ
coordinate as the counterclockwise angle in degrees with respect to θ
=0°
;
a saturation and hue value conversion subroutine in which the r, θ
coordinates are converted to saturation and hue values by projecting the RGB coordinates of the three primary colors RED, GREEN, and BLUE onto said plane, determining the r, θ
coordinates of the projected points of the three primary colors, determining scaling factors for the respective primary colors and an offset for one primary color such as will provide a correlation of the r, θ
coordinates of the primary colors with the known saturation and hue values for the primary colors, and applying said scaling factors and offset to the r, θ
coordinates of the projected image pixel points to obtain the correct saturation and hue values, respectively, of the projected image pixel points,a non-highlight value coordinate determination subroutine, in which the value coordinate V of each valid pixel is determined, and a true RGB coordinate estimation subroutine, in which the true RGB coordinates of each pixel are estimated from said S, H, and V values of each valid pixel. - View Dependent Claims (26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50)
-
Specification