Improvement of satellite imagery based on the known targets.

Understanding World Food Economy with Satellite Images and AI. It has become possible to observe the food growing from satellites daily at a global scale. Using it we can understand agriculture-specific insights and predict productivity of the commodity crops like corn, wheat and sugarcane . This talk will help to understand current publicly available satellite imagery data, how to inject it into the data pipeline and how to train and deploy AI/ML models based on it. With the active launches of new government and commercial Earth Observation satellites it has become possible to observe the food growing process from satellites daily at a global scale for the commodity crops like corn, wheat and sugarcane. Based on the satellite data we can identify and share agriculture-specific insights like: presence of farming activity, presence of irrigation systems, crop classification and productivity assessment. A pipeline starts with a set of images specifically designed for daily monitoring the growth of commodity crops: corn, soybean, rice and wheat. To process this data we use our processing and delivery system with AI/ML (boosting) used for understanding vegetation patterns and AI for scaling the models on other climate zones. The main challenges of this approach is availability of regional dataset to understand the context and working with massive datasets like over 10PB of just Open Source Earth Observation data.

Who is this presentation for?
Researchers, developers and decision-makers in data analytics who want to use geospatial context in their work for agriculture and soft commodities prices prediction

Prerequisite knowledge:
Very basic image processing, basic knowledge about geospatial and Earth observation satellites

What you'll learn?
An overview of existing data sources for analyzing food production patterns at global scale, overview of implementation techniques to process data cubes covering the whole World with various remote sensing data source and a case study of how to iterate with data processing and delivery infrastructure for satellite imagery.

MS in Computer Science from MIPT and Skoltech (Russia) and MIT (US). Head of Data at Astro Digital - Satellite Mission as a Service company. I’ve designed and developed the high-level product vision, data processing infrastructure and use of data workflows. Previously I was a co-founder and CTO of ImageAiry - recommendation system of satellite data services. I designed business and technical logic of the recommendation system. I've spoke about Open Source satellite imagery at Scale By The Bay (2017,2018), Space Dent (2016) and Strata Data (2019) Won NGA's Expedition Competition with the image detection tool for medium-resolution data in 2016.