Many big-box retailers like Walmart, Target, and Costco are increasingly under pressure to provide near real-time insights into their consumer-base while at the same time maximizing profits through the exploitation of Big Data. Investors similarly struggle with this same problem as they try to anticipate trends and bet on the stores and locations that will thrive in the years ahead.

Satellite imagery is being used as a way to understand consumer trends and meet these challenges. However, manually looking at hundreds, thousands, or even millions of images in order to gain actionable intelligence is difficult and time consuming.

Solution

The ENVI Deep Learning module can be used to take imagery and create alerting and monitoring services that help stores and investors save money and help predict future trends. One way the ENVI Deep Learning module has been applied to understand consumer spending is by counting the number of cars in store parking lots.

Science has been baked into the ENVI Deep Learning module so the user only has to put point and line features on each car. They don’t have to worry about the spatial and spectral characteristics – the tools within the product take that into account and adjust parameters automatically.

Since the ENVI Deep Learning module was specifically designed to work with remotely sensed imagery to solve geospatial problems, it is perfectly suited for this application. Users have intuitive tools and workflows that don’t require programming and can easily label data and generate models with the click of a button

Extracting consumer information from imagery in this automated way allows for timely, low manpower macro and microscopic insights. While traditional image analysis can be used for this type of task, ENVI Deep Learning excels because of its ease of use and powerful functionality that can be utilized with just a few clicks of the mouse.

Benefits

Since the ENVI Deep Learning module can quickly and accurately count cars in large quantities of images, it can be used to gain near real-time information and evaluate historical trends. It can help reliably predict sales and spot abnormalities. Additionally, its seamless integration with ENVI enables robust data preprocessing and more accurate results.

If you have a problem you think can be solved using deep learning, reach out to us at GeospatialInfo@NV5.com and one of our experts will discuss it with you.