Abstract: A block-base error measure method for object segmentation includes the steps of dividing a reference image having an object into plural non-overlapping blocks, superimposing the reference image with a segmented image to obtain an error ratio of the block to define an enhanced equation and a modification equation to suppress the scattered error and enhance the contribution of the region error, so as to calculate the error amount of the segmented image and evaluate the performance of image segmentation. Compared with the conventional error measure method based on pixels, the present invention provides a more accurate high-level semantic.

Abstract: A block-base error measure method for object segmentation includes the steps of dividing a reference image having an object into plural non-overlapping blocks, superimposing the reference image with a segmented image to obtain an error ratio of the block to define an enhanced equation and a modification equation to suppress the scattered error and enhance the contribution of the region error, so as to calculate the error amount of the segmented image and evaluate the performance of image segmentation. Compared with the conventional error measure method based on pixels, the present invention provides a more accurate high-level semantic.

Abstract: A method of extracting and reconstructing court lines includes the steps of binarizing a court image of a court including court lines to form a binary image; performing horizontal projection for the binary image; searching for plural corners in the binary image and defining a court line range by the corners; forming plural linear segments from images within the court line range by linear transformation; defining at least one first cluster and at least one second cluster according to the characteristics of the linear segments and categorizing the linear segments into plural groups; taking an average of each group as a standard court line and creating a linear equation of the standard court line to locate the point of intersection of the standard court lines; and reconstructing the court lines according to the point of intersection.

Abstract: A method for creating a face replacement database includes steps of creating a face database for storing a plurality of replaced images with a face image rotation angle by using a method for estimating a 3D vector angle from a 2D face image, and defining a region to be replaced in the replaced image. The method for estimating a 3D vector angle from a 2D face image includes the steps of creating a feature vector template; detecting a corner of eye and mouth in a face image; defining a sharp point in a vertical direction of the quadrilateral plane, and converting the vertices into 3D coordinates; computing the four vectors from the sharp point to the four vertices to obtain a vector set, and matching the vector set with the feature vector model to obtain an angle which is defined as a rotation angle of the input face image.

Abstract: A method for automatic video face replacement includes steps of capturing a face image, detecting a rotation angle of the face image, defining a region to be replaced in the face image, and pasting a region to be replaced of one of the replaced images having the corresponding rotation angle of the face image into a target replacing region. Therefore, the region to be replaced of a static or dynamic face image can be replaced by a replaced image quickly by a single camera without requiring a manual setting of the feature points of a target image. These methods support face replacement at different angles and compensate the color difference to provide a natural look of the replaced image.

Abstract: A method of extracting and reconstructing court lines includes the steps of binarizing a court image of a court including court lines to form a binary image; performing horizontal projection for the binary image; searching for plural corners in the binary image and defining a court line range by the corners; forming plural linear segments from images within the court line range by linear transformation; defining at least one first cluster and at least one second cluster according to the characteristics of the linear segments and categorizing the linear segments into plural groups; taking an average of each group as a standard court line and creating a linear equation of the standard court line to locate the point of intersection of the standard court lines; and reconstructing the court lines according to the point of intersection.

Abstract: A method for estimating a 3D vector angle from a 2D face image, a method for creating a face replacement database and a method for replacing a face image includes steps of capturing a face image, detecting a rotation angle of the face image, defining a region to be replaced in the face image, creating a face database for storing replaced images corresponding to the region to be replaced, and pasting one of the replaced images having the corresponding rotation angle of the face image into a target replacing region. Therefore, the region to be replaced of a static or dynamic face image can be replaced by a replaced image quickly by a single camera without requiring a manual setting of the feature points of a target image. These methods support face replacement at different angles and compensate the color difference to provide a natural look of the replaced image.

Abstract: A method for automatic video face replacement includes steps of capturing a face image, detecting a rotation angle of the face image, defining a region to be replaced in the face image, and pasting a region to be replaced of one of the replaced images having the corresponding rotation angle of the face image into a target replacing region. Therefore, the region to be replaced of a static or dynamic face image can be replaced by a replaced image quickly by a single camera without requiring a manual setting of the feature points of a target image. These methods support face replacement at different angles and compensate the color difference to provide a natural look of the replaced image.

Abstract: A method for creating a face replacement database includes steps of creating a face database for storing a plurality of replaced images with a face image rotation angle by using a method for estimating a 3D vector angle from a 2D face image, and defining a region to be replaced in the replaced image. The method for estimating a 3D vector angle from a 2D face image includes the steps of creating a feature vector template; detecting a corner of eye and mouth in a face image; defining a sharp point in a vertical direction of the quadrilateral plane, and converting the vertices into 3D coordinates; computing the four vectors from the sharp point to the four vertices to obtain a vector set, and matching the vector set with the feature vector model to obtain an angle which is defined as a rotation angle of the input face image.

Abstract: A method for estimating a 3D vector angle from a 2D face image, a method for creating a face replacement database and a method for replacing a face image includes steps of capturing a face image, detecting a rotation angle of the face image, defining a region to be replaced in the face image, creating a face database for storing replaced images corresponding to the region to be replaced, and pasting one of the replaced images having the corresponding rotation angle of the face image into a target replacing region. Therefore, the region to be replaced of a static or dynamic face image can be replaced by a replaced image quickly by a single camera without requiring a manual setting of the feature points of a target image. These methods support face replacement at different angles and compensate the color difference to provide a natural look of the replaced image.

Abstract: A visual object tracking method includes the steps of: setting an object window having a target in a video image; defining a search window greater than the object window; analyzing an image pixel of the object window to generate a color histogram for defining a color filter which includes a dominant color characteristic of the target; using the color filter to generate an object template and a dominant color map in the object window and the search window respectively, the object template including a shape characteristic of the target, the dominant color map including at least one candidate block; comparing the similarity between the object template and the candidate block to obtain a probability distribution map, and using the probability distribution map to compute the mass center of the target. The method generates the probability map by the color and shape characteristics to compute the mass center.

Abstract: A visual object tracking method includes the steps of: setting an object window having a target in a video image; defining a search window greater than the object window; analyzing an image pixel of the object window to generate a color histogram for defining a color filter which includes a dominant color characteristic of the target; using the color filter to generate an object template and a dominant color map in the object window and the search window respectively, the object template including a shape characteristic of the target, the dominant color map including at least one candidate block; comparing the similarity between the object template and the candidate block to obtain a probability distribution map, and using the probability distribution map to compute the mass center of the target. The method generates the probability map by the color and shape characteristics to compute the mass center.

Abstract: Disclosed are virtual glasses try-on method and apparatus, and the apparatus includes an image capturing unit for capturing a user's image and a processing device for detecting a face image from the user's image, storing a glasses model, defining a first feature point of a lens of the glasses model and a second feature point at the center of a frame, and obtaining vertical vectors of the first and second feature points to find a third feature point. Two eye images are searched and binarized into a binarized picture that is divided into an eye area and a non-eye area. A center point between first and second extreme values is found, and vertical vectors of the first extreme value and the center point are obtained to find an example point. An affine transformation of the feature points is performed and attached to the face image to form a try-on image.

Abstract: Disclosed are virtual glasses try-on method and apparatus, and the apparatus includes an image capturing unit for capturing a user's image and a processing device for detecting a face image from the user's image, storing a glasses model, defining a first feature point of a lens of the glasses model and a second feature point at the center of a frame, and obtaining vertical vectors of the first and second feature points to find a third feature point. Two eye images are searched and binarized into a binarized picture that is divided into an eye area and a non-eye area. A center point between first and second extreme values is found, and vertical vectors of the first extreme value and the center point are obtained to find an example point. An affine transformation of the feature points is performed and attached to the face image to form a try-on image.