Hughes Electronics Corporation

Michael W Autry

William B Noble

Leo H Hui

Thomas W Bradley

A range cell formation process that are used to enhance the quality of an image following a perspective transformation from range-azimuth coordinates (B-scan) into elevation-azimuth coordinates (C-scan). The present range cell formation process eliminate situations where there are missing data points at the far range, and large areas that have the same data point at the near range. The range cell formation process evaluates the data content in range cells on a case by case basis using a prioritized system to determine what is to be displayed so that a high level of image contrast is maintained. In the far range, range cell data that are in between range points defined by every two adjacent elevation points are processed and a priority system is used to determine the best intensity value to use for the elevation points. The priority system is such that bright objects have the highest priority, followed by dark objects, and then followed by an average background level. Range cells are processed to determine which priority is met, and then the appropriate intensity is used. If a bright object is detected in the range cells, then the maximum intensity of the range cells is used. If a dark object is detected, then the lowest intensity of the range cells is used. Otherwise, the average intensity of the range cells is used. In the near range, no enhancements are made to reduce the size of large areas, but the size of bright objects is reduced, thereby providing a sense of increased focus in the near range.

Range cell formation processing

Integrity checking apparatus and procedures for checking safety-critical high throughput data and image transformations. The present invention randomly samples input and output data streams and uses the sampled data in either a forward or backward transformation to verify that the input data and output data are consistent. Because only a small fraction of the overall data is examined, the computational workload is very small by comparison, and processing may be more efficiently implemented. More particularly, in order to check the data, input data is transformed into output data, the output data is sampled, the input data is sampled corresponding to the sampled output data, and the input data is calculated from the output data, using forward or backward (inverse) data transformations. If the sampled input data and sampled output data correspond, then the output data is deemed to be correct. If the data does not correspond, then an error signal is generated.

Growing community of inventors

Woodland Hills, CA, United States of America

Michael W Autry