{"id":1096531,"date":"2021-02-22T14:39:14","date_gmt":"2021-02-22T22:39:14","guid":{"rendered":"https:\/\/www.esri.com\/arcgis-blog\/?post_type=blog&p=1096531"},"modified":"2021-02-23T15:50:56","modified_gmt":"2021-02-23T23:50:56","slug":"interactive-on-demand-object-detection-in-arcgis-pro","status":"publish","type":"blog","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/3d-gis\/interactive-on-demand-object-detection-in-arcgis-pro","title":{"rendered":"Interactive on-demand object detection in ArcGIS Pro"},"author":10962,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_acf_changed":false,"_searchwp_excluded":""},"categories":[23771,22941],"tags":[],"industry":[],"product":[36561],"class_list":["post-1096531","blog","type-blog","status-publish","format-standard","hentry","category-3d-gis","category-mapping","product-arcgis-pro"],"acf":{"short_description":"The Interactive Object Detection tool which detects objects by analyzing data in your map, is new functionality in ArcGIS Pro 2.7.","flexible_content":[{"acf_fc_layout":"content","content":"

Detecting objects in a map or scene just got easier! ArcGIS Pro 2.7 contains new functionality to detect objects by analyzing data in your map via the newest member of the exploratory analysis tool family \u2014 Interactive Object Detection.<\/p>\n

Data acquisition is a common obstacle for many GIS tasks. You don\u2019t always have the feature classes you need. The Interactive Object Detection tool helps provide data by finding and extracting objects of interest \u2014 such as windows and doors \u2014from data, like 3D objects, imagery, and even integrated mesh scene layers<\/a> that are already displayed in the view by using object detection techniques<\/a>.\u00a0 The detected objects are saved as point features that can then be used as inputs for other analysis tools.<\/p>\n

This Interactive Object Detection tool is different than other exploratory analysis<\/a> tools in that it uses a trained deep learning<\/a> model.\u00a0 Don\u2019t let the phrase deep learning<\/em> scare you! I was unfamiliar with it at first too, but after learning the terminology and practicing with the tool, I was totally wowed.<\/p>\n

For the rest of this blog, I\u2019ll walk you through the basics of the Interactive Object Detection tool, showing you how to extract windows and doors from building layers using a pre-trained model. Soon, you\u2019ll be finding more models and detecting more objects like a pro.<\/p>\n

Before you get started, you need to install the Deep Learning libraries<\/a> and have an Advanced license level<\/a>.<\/p>\n

To use the tool, open a 3D scene with layers for textured buildings in ArcGIS Pro 2.7.\u00a0 I found the “Tourism in Lyon” web scene from ArcGIS Living Atlas of the World works well.\u00a0 In ArcGIS Pro, click the Analysis<\/strong> tab and expand the Exploratory 3D Analysis<\/strong> gallery and choose Object Detection<\/strong> to open the Exploratory Analysis<\/strong> pane for the tool.<\/p>\n

 <\/p>\n"},{"acf_fc_layout":"image","image":{"ID":1096841,"id":1096841,"title":"ObjectDetection01_pane","filename":"ObjectDetection01_pane.png","filesize":9930,"url":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","link":"https:\/\/www.esri.com\/arcgis-blog\/products\/arcgis-pro\/3d-gis\/interactive-on-demand-object-detection-in-arcgis-pro\/objectdetection01_pane","alt":"","author":"10962","description":"","caption":"The Exploratory Analysis pane for the Object Detection tool with default settings before downloading the Esri Windows and Doors model. ","name":"objectdetection01_pane","status":"inherit","uploaded_to":1096531,"date":"2020-12-23 00:31:09","modified":"2020-12-23 05:32:24","menu_order":0,"mime_type":"image\/png","type":"image","subtype":"png","icon":"https:\/\/www.esri.com\/arcgis-blog\/wp-includes\/images\/media\/default.png","width":272,"height":695,"sizes":{"thumbnail":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane-213x200.png","thumbnail-width":213,"thumbnail-height":200,"medium":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","medium-width":102,"medium-height":261,"medium_large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","medium_large-width":272,"medium_large-height":695,"large":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","large-width":272,"large-height":695,"1536x1536":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","1536x1536-width":272,"1536x1536-height":695,"2048x2048":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","2048x2048-width":272,"2048x2048-height":695,"card_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane-182x465.png","card_image-width":182,"card_image-height":465,"wide_image":"https:\/\/www.esri.com\/arcgis-blog\/app\/uploads\/2020\/12\/ObjectDetection01_pane.png","wide_image-width":272,"wide_image-height":695}},"image_position":"center","orientation":"horizontal","hyperlink":""},{"acf_fc_layout":"content","content":"

Set-up overview<\/h2>\n"},{"acf_fc_layout":"content","content":"

There are two categories of parameters to set: one for how the model behaves, the other for the output feature layer. To define the model, click the Model<\/strong> drop-down and click Download<\/strong> to automatically get the Esri provided pre-trained model for windows and doors. Select Windows<\/strong> from the Classes<\/strong> drop-down and define the other model parameters<\/a> as you see fit. I kept the defaults.<\/p>\n

For the output feature layer, provide a name, an optional description to be included in the attribute table, and choose the returned shape.\u00a0 That depends on if I want the detected object marked with an X at the center point (\u201cLocation Point\u201d), or see the object displayed as a semi-transparent filled bounding box (\u201cVertical Bounding Box, Horizontal Bounding Box)? \u00a0I used Vertical Bounding Box.<\/p>\n

Before running the tool, I want to share these two tips to improve your results:<\/p>\n