Essess’ team of machine vision engineers includes leading academics, machine learning engineers, and 3D geometric modeling experts. To analyze the over 6 terabytes of data collected by each vehicle each night, Essess has developed the most advanced proprietary pattern recognition software capable of identifying at high geospatial resolution physical assets as diverse as buildings, walls, doors, windows, soffits, attics, foundations, exterior surface materials, telephone and utility poles, electric grid distribution assets like transformers, primary and secondary wires, street lights, road network assets, and much more.
Essess’ algorithms span classical computer vision methods like linear classifiers, deep learning and convolutional neural networks, unsupervised learning techniques, and geometric modeling in both the 2D and 3D spaces. Moreover, Essess’ capabilities take advantage of the combination of optical, near infrared, long wave infrared and 3D sensors that make our data collection and analysis entirely unique within the field of mobile mapping. Our machine vision capabilities enable us to automate the process of identifying, classifying and analyzing objects across large physical environments and has applications across building efficiency, city and utility infrastructure, oil & gas assets, road network mapping, and nighttime navigation.