I always believed that computer vision has potential to contribute very strongly in general computing. This became more obvious in the lst few years. Computing progressed significantly and digital cameras started appearing and used every where. Computer vision as a research field has been very popular. Academic conferences attract lots of people and many active research groups are very active in this area in all parts of research world. However, computer vision has not yet become main stream in computing.
On Friday and Saturday of this past week, I had a chance to visit two companies in the Bay area: Appscio and Cradle. Both are very interesting and are relevant to computer vision becoming main stream. Appscio believes that using computer vision metadata related to objects and events should be extracted and then meta data processing environment could be used to combine this data with other sources. I like this approach. Cradle uses DSP to do video processing in real time to extract important information that could be used by other systems. They procide programming environments to design algorithms to extract required information, if you kow how to, from sources (upto 16 video cameras) in real time multiple. These two companies are interesting because they clearly believe that visual processing of data from cameras is becoming the mainstream task and are developing real systems based on their belief.
This is very encouraging to me. I can use both of these systems in my own research group.