Abstract: The present invention relates to computer production of images. Three-dimensional graphics data (D3D) is automatically rendered by means of a GPU (330), which is adapted to receive two-dimensional image data. This data contains a number of image points, which each is associated with color information (r, g, b), transparency information (a), and depth buffer data (Z) that for each of the image points specifies a distance between a projection plane and a point of a reproduced object in the scene. A buffer unit (320) storing the image data is directly accessible by the GPU (330). The GPU (330), in turn, includes a texture module (331), a vertex module (332) and a fragment module (333). The texture module (331) receives the color information (r, g, b) and based thereon generates texture data (T) for at least one synthetic object in the synthetic three-dimensional model (V).

Abstract: The present invention relates to automatic modeling of a physical scene. At least two images (I1, I2) of the scene are received, which are taken from different angles and/or positions. A matching module (130) matches image objects in the first image (I1) against image objects in the second image (I2), by first loading pixel values for at least one first portion of the first image (I1) into an artificial neural network (133). Then, the artificial neural network (133) scans the second image (I2) in search of pixels representing a respective second portion corresponding to each of the at least one first portion; determines a position of the respective second portion upon fulfillment of a match criterion; and produces a representative matching result (M12). Based on the matching result (M12), a first calculation module (140) calculates a fundamental matrix (F12), which defines a relationship between the first and second images (I1, I2).