Research Article
Efficient Parallel Implementation of Active Appearance Model Fitting Algorithm on GPU
Algorithm 2
CPU-based method for computing
.
// c: the appearance parameters. | // and : the matrices describing the modes of | // variation derived from the training set. | // s: the shape in the model. | // : the mean shape. | // ST: the similarity transformation. | // S: the shape in the image. | // im: the input image. | // SP: samples textures from image. | // : the image texture. | // : the mean texture. | // ZMUL: zero-mean normalization. | // Align: align to the mean texture. | // : the model texture. | // r: the residual texture. | // E: the error between the model and image. | () + c; | () ST(s); | () SP(im, S); | () ZMUL(); | () Align(); | () + c; | () ; | () ; |
|