
Explore artificial intelligence and deep learning applied to geographic information systems with ArcGIS Pro, from theory to practical use of pre-trained models for car detection, crowd counting, and vegetation analysis.
Explore how artificial intelligence underpins deep learning in ArcGIS Pro, using neural networks to analyze raw data and detect patterns, and compare it with traditional machine learning.
Explore geo AI, merging artificial intelligence with geographic information systems, and apply deep learning for image analysis using convolutional neural networks.
Apply deep learning in ArcGIS Pro to perform image classification, object detection, semantic segmentation, and instance segmentation, streamlining geoprocessing and automating workflows for building footprints and land use datasets.
Train a deep learning model in ArcGIS Pro through stages: prepare training data, train the model, run it, and review results, producing a reusable package for object detection and classification.
Discover Esri's pre-trained models in ArcGIS Pro, downloadable from ArcGIS Living Atlas and ready to use without data preparation or training, with adjustable parameters for task-specific workflows.
Ensure smooth image analysis with ai and deep learning in ArcGIS Pro by meeting hardware requirements: Ryzen 7 or Intel i7, 16gb ram, dedicated gpu (RTX 3050 ti), and ssd.
Download the matching deep learning libraries for ArcGIS Pro from GitHub, verify the Pro version in settings, extract the zip, and run the Pro deep learning installer to enable tools.
Explore image analysis tools in ArcGIS Pro, including deep learning tools for object and pixel classification, change detection, and object detection, with guidance on hardware requirements for mid-range machines.
Esri's pre-trained image analysis models in ArcGIS Living Atlas, including delineating agricultural fields, extracting building footprints, detecting elephants, and classifying land cover with Landsat eight and Sentinel two imagery.
Download six practice materials for car detection, counting people, extracting building footprints, plant disease detection, tree detection, and training a deep learning model on a one-centimeter orthophoto, with step-by-step guidance.
Detect cars in high-resolution aerial images with a pre-trained deep learning model in ArcGIS Pro, producing a car detection feature class for traffic planning, parking utilization, and urban analysis.
Automate crowd counting with a deep learning model in ArcGIS Pro, using a 2000 x 2000 jpg with 3 RGB bands to output a feature layer of detected people.
Extract building footprints from high resolution imagery in ArcGIS Pro using a pre-trained deep learning model, automate digitization with geo AI tools, and output footprint feature class for urban planning.
Learn to detect plant diseases in fruit and vegetable plants using a pre-trained deep learning model in ArcGIS Pro, classifying leaf images with 224x224 RGB inputs and GPU processing.
Apply a pre-trained deep learning model in ArcGIS Pro to detect trees in drone and high-resolution RGB imagery, producing a tree feature class with detection confidences (97 trees identified).
Train an object detection model in ArcGIS Pro to automatically identify frailejon plants from orthophotos, enabling ecosystem conservation, water regulation monitoring, and long-term growth monitoring.
Prepare training data for a deep learning model in ArcGIS Pro by digitizing frailejones on an orthophoto, creating a frailejones shapefile, and exporting annotated training data with a 0.25 buffer.
Train the model by using the trained deep learning model tool with the training folder, save results to the model folder, using Retinanet for frailejones plant detection in high-resolution imagery.
Run the deep learning object detection model to identify frailejones in an orthophoto, load the model, configure output, enable non-maximum suppression, and review results on GPU.
Compare object detection with and without non-maximum suppression in ArcGIS Pro, run on GPU, then remove duplicates via a Python centroid-distance script to yield 13,256 unique frailejones.
Celebrate completing image analysis using artificial intelligence and deep learning in ArcGIS Pro by applying theory and practical exercises to your geospatial projects.
Welcome to the course Image Analysis Using IA and Deep Learning in ArcGIS Pro!
In this course, you will dive into the exciting world of artificial intelligence (AI) and deep learning, applied to geographic information systems (GIS). Throughout this journey, we will explore how AI is transforming geospatial analysis, providing us with powerful tools to solve complex problems across various industries, from car detection and crowd counting to vegetation analysis and urban planning.
The course is divided into two parts:
Theoretical Part: We will lay the foundation to understand the fundamental concepts of AI, Machine Learning, and Deep Learning. Additionally, we will see how these integrate with GIS systems and explore some of the most impactful applications of Deep Learning in this field. We will also examine the workflow needed to develop your own models and review the pre-trained models offered by ESRI, giving you a clear idea of their utility in spatial analysis.
Practical Part: We will move from theory to action with a series of hands-on exercises. We will begin with five examples of pre-trained models that will allow you to familiarize yourself with the power of these tools for tasks like car detection, crowd counting, construction footprint digitization, and plant disease identification. Finally, we will wrap up with an exciting project where you will create your own Deep Learning model to identify frailejón plants in a high-resolution orthophoto, using the advanced tools of ArcGIS Pro.
This course is designed for both beginners and those with prior experience to make the most of it. You will not only learn technical concepts but also gain practical skills that you can apply to your geospatial analysis projects.
Thank you for joining this adventure! I am confident that by the end, you will feel more empowered to tackle the challenges of spatial analysis with AI and Deep Learning. Let’s get started!