HStreaming today unveils real-time video content analysis for Big Data. The software allows customers to process, analyze, visualize, correlate, and react to thousands of concurrent video streams. HStreaming’s platform leverages Hadoop MapReduce for full parallelization of video processing and HStreaming’s proprietary engine for real-time delivery.
HStreaming takes a novel approach to video content analysis by viewing it holistically as a Big Data challenge. Video data is processed on the same platform with all other data types. HStreaming’s innovative highly-parallelizable low-latency MapReduce engine is fully integrated into the Hadoop ecosystem. That technology innovation brings the following advantages:
Analysts can concurrently analyze thousands of video streams in real time and combine video data with other structured and unstructured data sources such as text or transactional data. This approach is fundamentally different to traditional video analytics solutions which solely focus on video data.
HStreaming’s video analytics allows for analyzing video streams using a common query language. This enables to use the same rich set of analytical functions as for structured data like grouping, filtering, projections, aggregations, pattern matching, and clustering to results of video analysis, such as identified objects and faces.
HStreaming’s video analytics solution is available for both real-time analytics and batch processing. It can be used for training, discovery, clustering, and search over large sets of archived video footage. Analysts can also use many of the other Hadoop ecosystem products.
“By integrating video streams into a larger analytics platform with other data types, we enable organizations to focus on asking the right questions rather to think about tooling and integration. We treat all data sources equal and that gives analysts unprecedented flexibility to where to get insights from.”, says Jana Uhlig, CEO HStreaming.