OPTIMIZATION OF VISIBLE PIXELS FOR A SURROUND VIEW AND OTHER ENVIRONMENTAL VISUALIZATIONS

Abstract

In various examples, an ego machine can be equipped with any number and type of sensor(s), where the sensor(s) can be used to generate images of overlapping images, and where an inverse projection can be used to sample the image data from the corresponding sensor(s) and generate a view of the inverse projection through a specific or suitable viewport. Regional magnification factors or pixel densities can be measured or calculated, and where different sampling strategies can be used to sample the image data for different areas of the view of the inverse projection based on the regional magnification factors or pixel densities, in order to control local sampling strategies in local areas of the view of the inverse projection. Image quality processing can be performed on the pixels that are visible in the view of the inverse projection (e.g.,(instead of the original image data). The resulting visualization can be displayed on a screen visible to an occupant or operator of the ego machine.

Images