Teaching Modern Remote Sensing

all right let's start hello again everyone hope you are enjoying second day of the education summit so far welcome to this technical workshop on the topic of teaching modern remote sensing i'm cancerina cornea senior solution engineer with education team at esri some of you may know me already since i often go to university to do outreach before we start a little bit of housekeeping this station is being recorded and please mute your microphone and turn off your video camera during the presentation however please feel free to post questions or comments in the chat windows and we will answer as many as we can during the qa sessions and also before we continue we would like to ask you a couple poor questions kind of like an icebreaker in here the first one is getting to know you so i would like to ask my my colleague laura to open the first poll question we just want to know whether you are a faculty member or instructor researcher or student or others if there is none of them applied to you yeah so i'm going to give you like 10 seconds here to answer thank you it seems like it's coming let's give a little bit more time here awesome okay the polls close let's let's share the result here it's great to see like many of you the faculty member instructor because this is the this this session is actually designed for you but we welcome researcher we welcome student as well and others okay thank you we can close that result and i'm gonna ask um laura to help to open the second one so for uh all of you that are teaching are you actually currently teaching imagery and remote sensing now this is a multiple choices so you can you can choose more than one nice nice all right we can end the pond and then i also share the result awesome so about 42 of you teach the undergraduate introductory level and then 21 of you at the fence and the rest is actually um certified certificates program 10 and 20 percent of researcher and not teaching awesome okay so for those that teaching introductory level you're probably going to like the resources that we're going to share today thank you all right let's continue here so now i would like to introduce our presenter for today's session the first one is delfin khanna delvin the senior product engineer with learn arcgis team delphin has been working hard in designing and releasing the teaching material on introduction to remote sensing and introduction to imagery and remote sensing so this teaching material is designed to help the development of university level curricula on subject of imagery and remote sensing and this is available to all of you to access and adopt and use then we're gonna have emily window emily is the product manager with the arcgis imagery product team and emily will introduce the ready-to-use data apps and other resources to better enable you for teaching and research on imagery and remote sensing and at the end i will cover resources that we we covered and discussion and moderate decisions so we have a lot to cover so without further ado i'm going to stop staring and let delphine to take over therefore thank you sears yes thank you rina um sorry there's something unexpected but i'm going to be sharing my screen all right here we are good awesome all right so hi everyone and today i'm going to tell you about a learning package as rina mentioned that we created to aid the development of university level curricula on the subject of imagery and remote sensing first a few words about my team learn arcgis we develop learning materials based on real-world scenarios we work collaboratively with experts throughout esri as well as contributors external to ethereum and all our materials are available on our website in public access we created this package because we know that technology changes rapidly especially in the domain of imagery and remote sensing where there are constantly new tools and work workflows coming up we know that it is hard for you as educators to keep up so we wanted to help and we wanted our learning materials to demonstrate modern workflows and current best practices we also wanted them to be ready to use in the classroom or as assignments so that you instructors could save time we developed this project collaboratively with our colleagues from esri's imagery team as well as several other experts the results introduction to imagery and remote sensing offers a growing body of materials it is free to use for all audiences it is organized by themes the materials are introductory level and cover the basics of imagery and remote sensing but we also wanted to give a taste of engaging techniques such as working with drone imagery lidar point clouds and deep learning so there are three types of learning materials in the package interactive web apps hands-on labs and overview slides the interactive web apps are meant as short activities to illustrate a fundamental remote sensing concept such as special resolution or visualizing change you can use them in the classroom to demonstrate a concept or you can assign them as short homework hands-on labs are in-depth exercises that enable students to experience end-to-end workflows while gaining an understanding of how imagery and gis technology can solve real-world problems they take from 30 to 90 minutes to complete and you can assign them to your students in their original form or you can also adapt them to your teaching needs they are based on engaging scenarios such as measuring a shrinking lake in china tracking the development of the space canal over time exploring a volcanic eruption in hawaii and more they demonstrate a number of techniques like choosing band combinations and spectral indices unsupervised and supervised classification change analyses object detection with deep learning working with lidar point clouds drone imagery time series data etc so we very much hope that several of them will be of relevance to your course and we also have overview slides that cover fundamental imagery and remote sensing concepts they are offered in powerpoint and you can reuse them as desired to support your teaching like tech specific slides whatever you decide to do with them in terms of software we focus primarily on arcgis pro we also use arcgis online arcgis drawn to map for lessons relying on drone images as well as some web apps for the most introductory materials i will now give you a short demo so this is the website and this is the home page um you can see here that um it is subdivided in five sections which you can also find at the top here so if we look at the first section discovering imagery we are focusing here on understanding imagery fundamentals and you can see an example of a hands-on lab get started with imagery which exposes the students to various uses of imagery at this point this is very much exploratory you can see here some of the interactive web apps for instance on exploring imagery special resolution or exploring bands etc then on section two we have working with imagery where we focus on imagery workflows and start processing imagery uh for instance here we have a a hands-on lab on assessing burn scars with satellite imagery as well as preparing and rendering imagery with again more web apps um and labs um section three extracting information from imagery now we are starting to look into image interpretation image classification where we look at different types of classification for instance here calculating impervious surfaces from spectral imagery using a supervised classification and we also have examples of unsupervised classification a little bit lower in the analyzing change section so analyzing change looking at how things change over time we have the example here of classifying land cover to measure shrinking lake in china which is an example of unsupervised classification we also have examples with deep learning as well as looking at performing thematic accuracy assessment in the fourth section working with elevation and time now we are entering we're going beyond uh multispectral imagery uh looking at things like working with our digital elevation models lidar and 3d visualization working with drone data and multi-dimensional and temporal data and finally section 5 leveraging the power of gis we are sort of circling back that is showing that once you have uh processed your imagery and you have extracted information interesting information out of it you can now integrate this information those new layers into typical gis workflows such as this one which uses a suitability analysis to find suitable sites for sustainable shrimp farms in costa rica we also look at how you can share your results online for instance in this lab we create a web app that presents uh the effects of a mud slide with a sort of a before and after effect so i want to just show you a couple of um samples of the items we have uh here is the each section has an overview uh slide deck and i downloaded uh this one for you let me just yes here so this is how it looks so you know bunch of powerpoint slides here we have as an example what's the difference between raster vector and raster data um if we have a little bit uh let's see lower we have for instance here the effect of ground sample distance etc defining a number of essential concepts we also i want to show you an example of an interactive web app so we look at paraguay through time and this is an example where we so it's very simple we are learning about analyzing change we look at a specific example have a quick explanation of what this is about and then we have this very simple app which allows you to swipe and look at the difference between the paraguayan atlantic forest decoration in 1973 on the left and in 2000 on the right you can also um zoom in to look at the change in more details obviously we're seeing here that a lot of deforestation has happened as well as some strong changes in some of the water bodies uh due to uh dams and uh finally i'm going to show you um i just have to go back to my yes i'm going to show you an example of lab um so this is um yes we are here assessing hail damage in cornfields with satellite imagery so this is based on on a scenario all our lessons are based on scenarios so this is an actual event which happened where we actually um there was a hailstorm in alberta canada which had a strong impact on the cornfields and a lot of damage was done and so we are going to be uh analyzing this change and assessing the level of damage in various fields so you can see how our lesson lessons let me see if i can make it just a little bit bigger hold on here we are so we can see how um our lessons are step by step uh there are lots of illustrations so that the student can really orient themselves um and we give a number of explanations on how to do things and also why we are doing things what it means um you can see uh you know for instance examples here we are talking about a spectral signature for various types of um of uh cover and so to make things a little bit more lively i'm going to show you this workflow that the students perform in arcgis pro actually live so let's see here we are so when the students open the project they download the project with all the images we included and they open the project in pro and the first thing we are asking them to do is they see the two images before storm and after storm and they can swipe to look at the difference between the two and we can see that the top one is uh before storm and then we see here the one that is um after storm and we can see that yes something has happened but it's not really clear we can't really assess the change quite as um instantly um so we're going to of course do more uh what we are also going to do is the zoom interface is hiding part of the top i don't know how to get rid of this um okay let me see i might have to [Music] make this a little bit smaller i'm sorry because otherwise i literally cannot access sorry about that okay try to make it as big as possible nope okay sorry about this all right i think it will have to do like this okay so what we're going to do um next is the students will use the um let me first go out of swipe okay so the students will use the image information tool which is a really interesting uh tool to get a sense live dynamically of the values of the different bands that you can see on the right hand side of the screen right now and you can see the blue red sorry green and red bands as well as near infrared represented here in grey and we learned in the lesson that of course red and gray bands are essential to assess vegetation health we next are going to be computing a vegetation index on each image starting with the first one the before stone we do that by going to indices and we will choose the savvy index which is a well-known index for assisting health index that uses the red and near-infrared bands so all we need to do is provide the correspondence that is said that the near infrared band in our image is a fourth band and then the red in red band is the third band and here we get the result which is um a a raster where the white whiter areas correspond to the healthiest vegetation and then the darker ones correspond to either unhealthy vegetation dying vegetation or unvegetated areas we do the same for the second one so same thing fourth third all right and now that we have the two we're going to compute the difference between the two using a raster function all of this is done dynamically you see it's very quick very fast so we use the compute change raster function um indicate that we want to do the difference between the after uh savvy raster after storm and the before storm uh savvy raster and this is our result which aware the um the the darker purple um areas show that the fields that have been the most affected by this hailstorm and i don't have to show you the time to show you like the end the the rest but i will show you the final result where we um actually use do you by putting it at the top we use um zonal statistics as table to compute for each field of interest uh its average loss in vegetation and in healthy vegetation uh with the lower areas in yellow um going from the the lower um [Music] the low damage um in yellow to the highest damage in darkrai so this is um this was just a quick overview and um i hope you enjoyed it i think i am going to stop here because i have filled up my time so um yes um rina i think it's back to you i'm going to stop sharing perfect thank you delphine i think all of us can see that this teaching material have a lot to cover rich content slide decks interactive web apps and hands-on lab exercises and activities as well including the data set based on the real world scenario i know that as educators most of us have no time to really update our course materials regularly so this package can be a great help and you can pick and choose the lessons that you like to incorporate into the class you can customize you can actually also work with us such as if you want to contribute other content and lesson as well again this package is available now at no cost you can download it you can customize it you can inter interact with that but we are looking forward for any feedback from you so we can continue make it better however this package is not the only one that you can help you with your courses emily will also bring another additional resources that can help you for teaching so emily i miss yours great thanks nina um hi everyone thanks for joining us uh so i'm gonna spend some time uh talking about resources from esri that you can use to enrich and build on the imagery and remote sensing curriculum that delphine just talked about first i'm going to provide a really brief overview of imagery and arcgis just for some context so the imagery capability is really shorthand for a ton of functionality which we generally divide into five sub-capabilities which you can see here um so content uh refers to both the content that we support as well as online imagery terrain and gis layers that we provide from the living atlas we'll get a look at that in a second image management capabilities let you catalog publish and consume imagery efficiently and securely with image mapping you can create authoritative maps and products like orthomosaics dems or 3d mesh from the imagery collected by drones aerial sensors or satellites with image analysis arcgis provides advanced analytical tools to extract location-based information from imagery including image classification change detection deep learning and more and image visualization leverages human interpretation to extract information from imagery so that includes user experiences that are built for motion imagery or stereo intuitive search and discovery interfaces and tools for sharing and telling stories with your imagery and so the result is a really rich system for anything that users want to do with imagery um so that's how we understand and talk about the imagery capability in arcgis but it's also important to understand that when we're talking about arcgis that can refer to desktop enterprise and software as a service deployments so in a desktop environment software is installed and accessed on a personal computer in an enterprise environment your organization installs and manages the software on your own servers for the use of a lot of users simultaneously in a sas environment we host and manage the software and sometimes the data and you access it via the internet so here you can see that each of esri's flagship offerings in each environment for each of them there's an imagery extension so all of those extensions are sometimes referred to together as arcgis image which is a suite of products deployed to implement imagery capabilities and so how you implement those really depend on your organization all of those imagery sub capabilities we talked about are represented across all three environments and then we both also have focused imagery applications that stand on their own but can also be integrated with the rest of arcgis so as we start discussing imagery resources i'm going to focus on sas offerings for content and for telling stories with your imagery that you should be able to leverage as educators so let's take a look at some of those resources that you can use to expand your curriculum so we'll start with imagery and apps that are just ready to use out of the box so the arcgis living atlas of the world provides a really massive collection of ready-to-use authoritative data layers accessible via the internet curated and prepared for easy integration with arcgis so while there are a number of useful layers including feature data sets rasters of environmental variables all kinds of stuff i'm going to highlight the imagery resources so landsat satellites have been continuously gathering medium resolution multi-spectral images of the globe for almost 50 years esri's landsat service now contains more than a million individual scenes with new scenes added daily the multi-spectral landsat layer includes both the full landsat 8 imagery archive and historic gls epoch imagery which extends back to 1975. it also comes packaged with the collection of band combinations and indices that are rendered on demand some of which you saw how quickly that can work in delphine's demo and we'll see some in action with sentinel 2 in a second so currently this is analysis ready data with the top of atmospheric correction applied but we're in the process of releasing a new service comprised of landsat 8 collection 2 level 2 imagery which will provide you with atmospherically corrected data and look for that service in coming weeks so sentinel 2 is an earth observation mission operated by the european space agency since since 2015. esri sentinel 2 image service is updated with new 13-band imagery daily and includes all sentinel 2 scenes going back 14 months the imagery is analysis ready with top of atmosphere correction applied and similar to landsat it also includes a collection of band combinations and indices that can be visualized on the fly so the national agriculture imagery program typically produces 0.6 to 1 meter resolution for band imagery for about half the continental u.s each year so nape image service includes all nape imagery published since 2010 and with the attached rendering options it can also be really quickly visualized using a color infrared band combination or ndvi um the world imagery is a tile layer offering one meter or better satellite and aerial imagery in many parts of the world and lower resolution satellite imagery worldwide so perfect for an imagery based map living atlas also provides foundational elevation layers that support analysis and visualization across the platform so by default the service returns 32-bit floating point elevation values that are appropriate for analysis but it also includes rendering options to visualize hill shades slope aspect and more so these imagery resources are generally appropriate for both visualization and analysis though there are some restrictions on the scale of analysis they can be consumed via the internet in arcgis pro in the map viewer in arcgis enterprise and arcgis online or consumed and shared via web apps if you're interested i'd recommend you checking out the tutorial on using image services in arcgis pro that's linked at the bottom from the imagery workflows website so in addition to these services esri also provides a collection of apps that do a good job of demonstrating what's possible with these services and also serve as useful teaching tools on their own so let's go ahead and take a look at one of these apps the earth observation explorer uh featured in a story map so we're going to use this app to explore a recent volcanic eruption in hawaii so on december 20th 2020 the kilauea volcano erupted it spilled lava into the crater boiled the water lake away and replaced it with lava 700 feet deep so we'll use the earth observation explorer app to visualize and explore imagery of the eruption so we'll navigate to the halemaumau crater and then switch to sentinel-2 imagery and with a natural color band combination it can be difficult to really see what's happening in the image so we could create a custom band combination in the app as you can see here however in this case we're actually going to use a rendering that's attached to the service the shortwave infrared rendering which has dynamic range adjustment for better visualization so now let's take a look at the individual sentinel two scenes before and after the eruption on december 20th to see how the eruption progressed so the extent of the lava is really clear with that shortwave infrared rendering and as we're looking through these images it looks like the january 7th image looks relatively clear so we'll go ahead and examine that image more closely so first we'll look at some spectral profiles from the image so here if we click the lava we get a very distinctive spike in the short wave infrared reflectance that you can see here and if we click the surrounding land cover the spectral profile of the cooled lava seems most similar to the impervious surfaces found in urban areas so we can also look more closely at the spectral dynamics of the image with an interactive spectral scatter plot if we plot bands against each other we can identify relationships between these bands that correspond with different land cover and by circling groups of pixels on the plot we can see where they fall on the map so here areas with low near infrared reflectance and relatively high shortwave infrared for example appear to be molten lava and here these pixels appear to be smoke or clouds and areas with low near infrared reflectance and low shortwave infrared reflectance seem to be cooled lava around the center of the crater so when you take a look at the story map you can check out this app for yourself or grow directly to the app and be sure to check out the masking and change detection tools also all right so well those services and demo apps are really useful it can also be really powerful to create resources that are tailored to your own curriculum so story maps are a really great way to integrate images video gifs interactive maps and more to tell a story these aren't designed explicitly for imagery but they do offer a really powerful way to tell stories about imagery or a vehicle for incorporating your imagery to tell a powerful story so in addition to the example story map you just saw um the url links to a gallery of map examples that incorporate imagery in interesting ways so i'd encourage you to check that out next configurable apps for imagery are intended to allow you to create focused apps for exploring imagery for exploring and visualizing imagery layers so as delphine discussed earlier apps built using these templates are actually incorporated into the remote sensing curriculum if you find them useful it's not that difficult to build your own so there are three designed for imagery imagery viewer is the most basic with tools to explore imagery through time and space with a swipe tool to help you compare image mask is designed to interactively highlight areas of change between two images and a service or to highlight areas that meet a user set threshold for common indices so areas that are above 0.25 and dbi for example image visit lets you review a predetermined sequence of locations in imagery so you can review attributes or identify features we've used this for validating deep learning results for example so to build a configurable app you author a web map using map viewer classic share the web map as an app configure your app settings and then publish and share the app so let's build an app for exploring and visualizing how las vegas land cover has changed over the last 30 years so all configurable apps start with a formatted web map which you can then share as an app so we'll use the imagery viewer to create our web app here so once the app is created it'll take you to the express settings which you can use to configure the most important settings so in the about section we can add a details tool to help the user understand what the app is and how to get started in the interactivity section we'll configure a swipe tool so users can compare images and then in the theme and layout we'll rearrange the tools so that the app is configured the way formatted the way we want it so at this point i could switch to the full setup to access more advanced settings but i can also publish a perfectly serviceable app from here so the app we've created um which we'll show up in a second is designed to illustrate how las vegas land cover has changed between 1991 and 2020 using landsat imagery so you can compare dates using the swipe you can use bookmarks to explore areas of interest you can see the census tracts that i added and the pop-ups from the web mac and the web map can be used to add context we can change dates to explore change through time in the city so overall these apps allow you to create resources that focus on current events or areas of interest for your students or on a specific part of your curriculum that may not have been addressed with the existing tools so the resources i've discussed up until now are probably most useful for introduction level instruction but where should you start if you're interested in diving deeper into basic arcgis concepts or exploring more more advanced topics so feature extraction using deep learning or analyzing time series imagery the imagery workflow site is intended to be a jumping off point for common imagery workflows in arcgis so the site has three types of contents the workflows which aggregate resources from across esri to support common imagery tasks tutorials which provide step-by-step instructions and sample data to help you get started and best practices which offer deep dive information about optimizing your workflow for efficiency scalability usability different data types and more so let's take a look at the site so there are quick links at the top to the full gallery of workflows imagery tutorials best practices and a collection of open source imagery tools the workflows are organized by sub capability to help you find relevant content faster and you also see featured imagery workflows which are generally the newest or most up-to-date content so let's go ahead and check out the full gallery of workflows which are really the core of the site so you'll see that if i start searching for classification um two workflows appear uh we'll go ahead and check out the one for performing image classification uh each workflow explains the arcgis solution to a common imagery problem so in this case performing image classification it then links to helpful resources from across esri to help you solve it so with each workflow you'll find links to help to blogs story maps videos learn lessons trainings developer resources github repos information from tech support or from the azure community which used to be geonet and you'll also find links to the tutorials and best practices where relevant so it's really just a one stop shop for supplemental resources from esri organized by topic and that brings us to the end so i really hope you'll take advantage of ezra's resources to help supplement and customize your curriculum and i'll hand it back to reena now thank you emily that's really great so everyone the message in here is you get a lot of resources in here is the one that built by learn team and also by the product teams emily ready to use data sets apps um also the other content let me share my screen now so we do have a lot of resources and we would like to continue to help you to update your teaching material or to use it for your research so i list them all in here what i'm going to do is i'm going to copy and paste into the chat we still have about six minutes for the question and answer in here so i think um i would like to start with with scott scott um is i already like talked with you several times during the email about the teachings with the remote sensing so the question from you scott is i you teach the remote sensing with variety students some for whom do not have previous gs experience then and um and all the lab exercises assume pure experience with arcgis pro or are they structured that they can be completed student with no gis background uh delving would you like to elaborate more on your answer on that please yeah yeah i i did put a quick answer to that effect in the in the chat but yeah no so there is no prior gis experience required um and there is no uh pro experience required because we know that many students who study gis study remote sensing have not done gis before we um in fact at the beginning of this whole endeavor when we started this project we uh spoke with a number of um educators and they told us that very clearly so we made sure that this is not the case so we do explain the concepts as much as possible and then also because we have this step-by-step approach and all those screenshots it really helps the students be oriented in pro so they don't have to like know where to find such and such tool um so it should really work out for students who don't have much experience in gis or pro okay that's a good question as well here uh delphine will these lessons will be updated when the new version of pro is released yes we are working on it we uh it's always as you as you yourself know if you have uh workflows that you maintain it's uh always you know there are always new wonderful versions that come out of the software and so our team we do have people who are testers and we try to get uh to them as soon as possible and so our commitment is to keep them as updated as possible so that you don't have to do that that uh extra effort that you know working on a workflow and you're like oh things have changed a little bit uh with this new release of pro or other software awesome and also i believe the emily teams will always update the workflow what is new so i'm gonna want to also put the attention for you to the resources uh we will have a new mooc which is a very brand brand new one in imagery in action so you're going to be released in august 11 so uh the registration is open now if you want to uh to join you or your students would like to join so i'm excited i'm also registered for that and we also have the link there the last link that i give you in the chat is imagery and remote sensing e-learning guide laura borden's comprehensively put all the lists from essay academy learn all the lessons that you can also adopt as well so i copy and paste um that link to the resources and the chat there and what we want to do is actually we really want to build the community for all these educators that teach remote sensing because remote sensing is high in our priority as well so if you need any assistance if you like to talk more with us if you want to actually contribute to the lessons or would you like us to help you updating your curriculum or content feel free to contact me i put my email there um and and delphi and emily and all the team at s3 we work together so we're going to pull the subject matter expert for you as well all right is there any question that i missed delvin or emily or laura i think it was 52 minutes maybe i'll add one more thing which is that i i saw some questions about people um asking sort of you know whether we there were some questions at the end of our labs for you know for grading we do not have that uh we have um professors use our indicators use our labs in different ways some of them will ask maybe for the students to do a quick report or give some specific screenshots along the way some others will also ask the students to do the workflow once following our guidance and then we'll give a new set of imagery and ask the student to apply what they learn in our workflow onto the new imagery which is a great way to test whether the students really absorbed all the all that was explained in the lab yeah i like as well the answer from lisa that said that she has her student to submit the final imagery or map into the organization rts online so the professor can see or you probably can ask the student to create a story map based on the what they learn and um so they can apply it to another scenario so that's really nice all right um okay so uh roderick said as it always has to be about grading can can we not see the part we're learning of beautiful learning true true lately you know there's many different form of grading i guess you know it can be like tell me the story about your learning so very good all right i think we reached the top of the hour here uh thank you so much again um uh if you still have any question if you want to contact us i put my email there i can put as well my email and then delphian and if you and emily if you want to put yours you're welcome as well so as i said that we work together closely laura i written it back to you decision you

Channel: Esri Events

Views: 253

Rating: 5 out of 5

Keywords: Esri, ArcGIS, GIS, Esri Events, Geographic Information System, Education Summit

Id: ftmSIPjLgTk

Length: 45min 49sec (2749 seconds)

Published: Thu Sep 09 2021