{"id":1072321,"date":"2020-12-08T10:50:50","date_gmt":"2020-12-08T18:50:50","guid":{"rendered":"https:\/\/www.esri.com\/arcgis-blog\/?post_type=blog&p=1072321"},"modified":"2021-01-05T14:28:57","modified_gmt":"2021-01-05T22:28:57","slug":"whats-new-in-the-arcgis-image-analyst-extension-arcgis-pro-2-7","status":"publish","type":"blog","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/imagery\/whats-new-in-the-arcgis-image-analyst-extension-arcgis-pro-2-7","title":{"rendered":"What\u2019s coming in the ArcGIS Image Analyst extension (December)"},"author":8452,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_acf_changed":false,"_searchwp_excluded":""},"categories":[22931],"tags":[757671,564122,186132,445012,115262],"industry":[],"product":[36561],"class_list":["post-1072321","blog","type-blog","status-publish","format-standard","hentry","category-imagery","tag-arcgis-pro-2-7","tag-change-detection","tag-deep-learning","tag-image-analyst-extension","tag-imagery","product-arcgis-pro"],"acf":{"short_description":"Here are some of the highlights of the new capabilities you\u2019ll see in the ArcGIS Image Analyst extension for ArcGIS Pro 2.7 (December 2020)","flexible_content":[{"acf_fc_layout":"content","content":"

ArcGIS Image Analyst<\/a> is the ArcGIS Pro extension that provides tools for advanced image visualization, exploitation, and geospatial analysis. The latest release features change detection, expanded deep learning capabilities, enhanced support for multidimensional data, improved motion imagery capabilities, and much more.<\/p>\n

Below, take a tour of all the new imagery and remote sensing-related features we\u2019ve added in the ArcGIS Pro 2.7 (December) release to help expand and streamline your workflows.<\/p>\n

 <\/p>\n

Change detection<\/h1>\n

Creating change maps using imagery is one of the most fundamental applications in imagery and remote sensing<\/a>. Identify changes due to human activity, abrupt natural disturbances, or long-term climatological or environmental trends.<\/p>\n

Use change detection tools to determine the type, magnitude, and location of change by comparing two raster datasets or by analyzing a time series image cube containing data collected for one area at different times.<\/p>\n

The new\u00a0Change Detection Wizard<\/strong> provides guided workflows for\u00a0categorical change detection,\u00a0timeseries change detection, and\u00a0pixel value change detection.<\/p>\n"},{"acf_fc_layout":"image","image":{"ID":1073481,"id":1073481,"title":"Change detection wizard","filename":"ChangeDetectionWizard.jpg","filesize":43831,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/imagery\/whats-new-in-the-arcgis-image-analyst-extension-arcgis-pro-2-7\/changedetectionwizard","alt":"Change detection wizard","author":"8452","description":"","caption":"Change detection wizard","name":"changedetectionwizard","status":"inherit","uploaded_to":1072321,"date":"2020-12-04 06:58:47","modified":"2020-12-04 06:59:11","menu_order":0,"mime_type":"image\/jpeg","type":"image","subtype":"jpeg","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":925,"height":699,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard-213x200.jpg","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","medium-width":345,"medium-height":261,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","medium_large-width":768,"medium_large-height":580,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","large-width":925,"large-height":699,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","1536x1536-width":925,"1536x1536-height":699,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","2048x2048-width":925,"2048x2048-height":699,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard-615x465.jpg","card_image-width":615,"card_image-height":465,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ChangeDetectionWizard.jpg","wide_image-width":925,"wide_image-height":699}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"

Deep learning<\/h1>\n

Deep learning is transforming the speed at which you can gain intelligence from imagery. This release adds new model types, multispectral support for training, multilabel classification, and better accuracy reporting for object detection.<\/p>\n

New model types<\/h3>\n

In addition to the models previously supported in ArcGIS Pro 2.6, we\u2019ve added support for new model types to train AI models, including YOLOv3, FasterRCNN, and DeepLab.<\/p>\n

YOLOv3 and FasterRCNN both support object detection workflows. FasterRCNN is slower than models like SSD, but it can provide results with higher accuracy. YOLOv3 is a one-stage detector, so it\u2019s fast.<\/p>\n

DeepLab supports pixel classification models. It is found to be better than UNet on all benchmarks. It also takes into account the same amount of context with fewer filters than UNet, so it can achieve higher accuracy. It performs best when there are classes that are drastically different in dimensions (grasslands versus fields, for example).<\/p>\n

Multispectral support for training<\/h3>\n

Most earth observation sensors provide multiple bands to leverage. With this release, the restrictions on working with a three-band image are eliminated. All bands in a multispectral image can be used for feature extraction.<\/p>\n

Training multispectral data with sparse training samples (another enhancement) eliminates the need for classifying every pixel in a training sample image for land-cover classification (as an example). This leads to significant productivity improvements. Additionally, pixels that are not classified will be ignored during the training process. The support for multispectral imagery enables users to leverage the spectral responses of bands outside the visible range.<\/p>\n

Multilabel classification<\/h3>\n

Now you can easily extract multiple types of features without having to train a model multiple times. For example, in one pass, the system can classify houses, houses with pools, and houses with pools and solar panels\u2014no need to train multiple models and then perform additional analysis to join common features into a single GIS dataset.<\/p>\n

Better accuracy reporting for object detection<\/h3>\n

The new accuracy reporting tool calculates the accuracy of a deep learning model by comparing the detected objects from the\u00a0Detect Objects Using Deep Learning\u00a0tool to ground truth data.<\/p>\n"},{"acf_fc_layout":"image","image":{"ID":1073361,"id":1073361,"title":"Deep learning enhancements","filename":"deeplearning1.jpg","filesize":293396,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1.jpg","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/imagery\/whats-new-in-the-arcgis-image-analyst-extension-arcgis-pro-2-7\/deeplearning1","alt":"Deep learning enhancements","author":"8452","description":"Deep learning enhancements","caption":"Deep learning enhancements","name":"deeplearning1","status":"inherit","uploaded_to":1072321,"date":"2020-12-04 05:31:07","modified":"2020-12-04 05:39:19","menu_order":0,"mime_type":"image\/jpeg","type":"image","subtype":"jpeg","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":1660,"height":755,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1-213x200.jpg","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1.jpg","medium-width":464,"medium-height":211,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1.jpg","medium_large-width":768,"medium_large-height":349,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1.jpg","large-width":1660,"large-height":755,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1-1536x699.jpg","1536x1536-width":1536,"1536x1536-height":699,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1.jpg","2048x2048-width":1660,"2048x2048-height":755,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1-826x376.jpg","card_image-width":826,"card_image-height":376,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/deeplearning1.jpg","wide_image-width":1660,"wide_image-height":755}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"

Multidimensional management, processing, and analysis<\/h1>\n

For multidimensional analysis, you can now identify trends in your data using new statistical methods, merge and summarize categorical data in a time series of rasters and attach processing templates directly to raster datasets\u2014including multidimensional datasets.<\/p>\n

These are all the new and enhanced multidimensional tools and functions:<\/p>\n