This is one of the market “sectors” that has not been addressed by previous beam analysis systems. Multi-machine, multi-camera users can use BeamGage to throw enough hardware at a problem to get the job done. A modern video analysis system should be designed to take advantage of multiples CPU cores. In other words, the location on the network where video data is collected should be independent (or transparent) of the location where data is analyzed.įorth was the realization that computer processors are not going to scale up in clock cycle speed as quickly as they will begin to scale up in number of CPUs per chip. Third was the realization that sources should be location transparent. In other words, the data sources are interchangeable and the analyzers compute results on theoretical beams and real beams in the exact same way. Second was the realization that data sources (cameras, files, and synthetic beams) should be treated by the analyzer in a symmetric fashion, the analyzer not knowing how to distinguish between them. The system would validate against theoretical beams and adhere tightly to the ISO specification.
The first real breakthrough was recognizing the need for a beam analysis system that was self validating. The BeamGage effort was all about creating the “Next Generation Profiler” (and sneaking a much-needed video transport backbone passed the bean counters, don’t tell anyone!). The constant need to interface new camera technology with different types of analysis tools, forced us to realize that analyzers and sources must be able to evolve independently of one another and on top of a solid video transport backbone. The system is the result of lessons learned from several predecessor systems. It took years for the total vision of BeamGage to emerge.
The remaining beam analysis breakthroughs have mostly been functional. In terms of breakthroughs, there are many hidden in the software architecture, but they are exciting only for the ardent software geek.
0 kommentar(er)
