Image processing method and apparatus for detecting an edge of an object within an image
An image processing method and apparatus capable of easily detecting an edge of an object from an input image, in which the edge is detected using one step, without a pre-processing step, and a complicated trigonometric function is not used for gradient detection. The image processing method includes setting a window within an input image, analyzing the window to determine directions of edges of objects within the image included in the window, detecting edge information including the edge directions, and processing and outputting the window using the edge information.
1. An image processing method in an image processing apparatus, the method comprising:
setting, by the image processing apparatus, a window within an input image;
analyzing the window to determine edges of objects within an image included in the window;
calculating a value of directional projection of each direction of the window;
calculating values of the directional projections exhibiting a maximum difference among differences between the values of the directional projections obtained based on the directions of the window;
detecting directions of the edges using the values of the directional projections exhibiting the maximum difference;
generating edge information including the edge directions; and
processing and outputting the window using the edge information.
2. The image processing method of claim 1 , wherein each value of a directional projection of each direction is calculated using:
dir
x
,
y
,
z
=
∫
x
∫
y
∫
z
g
(
x
,
y
)
δ
(
cos
x
+
sin
y
-
z
)
where x and y are image coordinates of the window, and z is a distance from a center point of the window to x and y coordinates.
3. The image processing method of claim 1 , wherein the values of the directional projections exhibiting the maximum difference among the differences between the values of the directional projections obtained based on the directions are calculated using:
diff i =max(dir i,k −dir i,l )
where i is an x-axis coordinate of the window, and k and l are y-axis coordinates.
4. The image processing method of claim 3 , wherein detecting the directions of the edges using the values of the directional projections exhibiting the maximum difference is performed using:
dir
=
arg
max
p
(
(
diff
p
·
1
2
+
diff
p
+
diff
p
+
1
2
)
-
(
diff
p
+
3
2
+
diff
p
+
4
diff
p
+
5
2
)
)
where diff p−1 , diff p+1 , diff p+4 , and diff p+5 are projection values exhibiting the maximum difference, and p indicates an edge direction.
5. The image processing method of claim 1 , wherein the window is set to at least a 5×5 pixel size.
6. An image processing apparatus comprising:
a window setting unit for setting a window within an input image;
an edge information detecting unit for analyzing the window to determine edges of objects within an image included in the window, calculating a value of directional projection of each direction of the window, calculating values of the directional projections exhibiting a maximum difference among differences between the values of the directional projections obtained based on the directions of the window, detecting edge directions of the edges using the values of the directional projections exhibiting the maximum difference, and generating edge information including the edge directions; and
an image processing unit for processing and outputting the window using the edge information.
7. The image processing apparatus of claim 6 , wherein the edge information detecting unit calculates each value of a directional projection of each direction using:
dir
x
,
y
,
z
=
∫
x
∫
y
∫
z
g
(
x
,
y
)
δ
(
cos
x
+
sin
y
-
z
)
where x and y are image coordinates of the window, and z is a distance from a center point of the window to x and y coordinates.
8. The image processing apparatus of claim 7 , wherein the edge information detecting unit calculates the values of the directional projections exhibiting the maximum difference among the differences between the values of the directional projections obtained based on the directions of the window using:
diff i =max(dir i,k −dir i,l )
where i is an x-axis coordinate of the window, and k and l are y-axis coordinates.
9. The image processing apparatus of claim 8 , wherein the edge information detecting unit detects each edge direction using the values of the directional projections exhibiting the maximum difference using:
dir
=
arg
max
p
(
(
diff
p
·
1
2
+
diff
p
+
diff
p
+
1
2
)
-
(
diff
p
+
3
2
+
diff
p
+
4
diff
p
+
5
2
)
)
where diff p−1 , diff p+1 , diff p+4 , and diff p+5 are projection values exhibiting the maximum difference, and p indicates an edge direction.
10. The image processing apparatus of claim 6 , wherein the window setting unit sets the window to at least a 5×5 pixel size.