Pseudo three dimensional image generating apparatus

Abstract

The invention discloses a method for obtaining the attribute data for the pixels in a uniform phase space and apparatus therefore, wherein the method comprises the steps of, first obtaining the top of a triangle, then projecting the viewing space coordinates of each top to a screen space coordinates, and determining the pixels influenced by the triangles positioned in the screen space according to the screen space coordinates, for each of the pixels influenced by the triangles, calculating a plurality of center-of-mass factors located at the uniform phase space.

Description

technical field [0001] The present invention relates to a graphics processing technique, in particular to a technique for performing triangular interpolation in a homogeneous space. Background technique [0002] Polygon (triangle) interpolation is a computationally intensive task for hardware graphics devices, since it is performed for each pixel of each object on the screen. Each triangle has multiple attributes that need to be considered, such as depth, color, textures, etc.; the quality of the final image mainly depends on the accuracy of the above-mentioned interpolation work. In addition, the speed at which the overall graphics framework is produced also depends on the speed at which this interpolation works. The hardware used to perform this interpolation is a critical component in modern graphics chips, and is often a key factor in determining the chip's development budget. [0003] Currently, there are many triangular interpolation solutions implemented in chips wi...

AI Technical Summary

Problems solved by technology

The above three types of technologies are all calculated for each pixel, so if you want to process medium and large triangles, the amount of calculation required will increase significantly; in addition, the implementation of the above three types of technologies requires considerable dedicated hardware cooperation, and also That is, the hardware resources used by it cannot be shared for other computing tasks

[0006] Specifically, the disadvantages of the first type of technique described above include: a. additional computational work is required for the above step of projecting all parameters to screen space (i.e., dividing by W) to linearize them; b. Projection steps, interpolation steps and correction steps to recover (recover) the true value (true value); c.1 / W interpolation work requires an additional parameter delta to establish calculations, so there is a problem of accuracy; d. The division operation of the interpolated 1 / W_pix value returns the additional real parameter value in each pixel; and e. requires considerable dedicated hardware cooperation

[0007] Disadvantages of the second type of technique above include: a. also having the above-mentioned step of projecting all parameters to screen space (i.e. dividing by W) to linearize them and its required extra computational work; b. also having The above-mentioned projection step, interpolation step and correction step for restoring the true value; c. also having the above-mentioned step of restoring the additional true parameter value in each pixel by dividing by the interpolated 1 / W_pix value; and d. Considerable dedicated hardware is required, and the hardware used cannot be shared for other computing tasks

[0008] The disadvantages of the above-mentioned third type of technology include: a. The calculation of the pixel centroid (barycentric) must be completed in the pixel level first, and the calculation workload will increase exponentially in the case of triangles with multiple pixels; and b. Considerable dedicated hardware

Images