Satellite Image Recognition

Bringing Observational Data to Enterprise Analytics

Several dozen new satellites are coming online in the next 18 months offering 24-hour refresh rates for HD planetary imagery. However, Satellite analytics and interpretation has always been a very expensive manual process given the challenges of object detection.

With automated detection results finally surpassing human results in 2015 and the introduction of deep learning detect objects, satellite analytics is now becoming generally available to enterprise analytics. Blackstone eIT’s Object Recognition Server is making satellite analytics economical, scalable, reliable, and extendable.

Conclusion suggest to support decision making

A core goal of data analytics is to discover useful information that suggests conclusions to support decision making. By discovering patterns – predictive and prescriptive models can be created to optimize business practices. Observational data adds additional insights to optimize these models further, adds predictive capabilities, and even detects expectation deviations to take corrective actions. Over the next 18 months, several dozen commercial satellites are planned to be launched into orbit that will offer satellite imagery for commercial uses. This will open wide ranging opportunities to extend geospatial big data to existing traditional enterprise analytics.

While there is a growing effort from Silicon Valley to unlock the value of petabytes of new satellite imagery by perfecting “machine learning”, BlackStone eIT has a solution that works today. The BlackStone eIT ORS Server is a proprietary technology developed by BlackStone eIT to handle our customers unanswered needs including image cleaning, aging, and distortion handling. It is based on a state-of-the-art algorithm utilizing a deep learning methodology.

Automated Satellite Image Object Detection Applications

Change Detection

By recognizing buildings or crops, changes in these monitored objects can be detected to predict crop yields or urban growth over time

Object Counting

Can track the presence of objects in restricted areas or track cars in parking lots to predict profit

Data Analytics

Combine observational data with other data sources to conduct more complex analytics such as urban planning

Deviation Detection

Detect deviations from predictive models early to take proactive corrective actions

Processing Satellite Imagery

Blackstone eIT Object Recognition Server is part of our video analytics platform. It is a scalable cloud solution capable of securely receiving multiple image or a video feeds in parallel to process for object recognition. We offer image cleaning, aging, and distortion handling based on a state-of-the-art algorithm utilizing a deep learning methodology. This solution can be scaled to service the most demanding workloads.

The results can be customized by BlackStone eIT to fit your specific needs. Our AI and smart analytics experts can assist you in turning this observational data into smart insights that can be integrated into your smart analytics platform.