Introduction:

In this tutorial, you'll learn how to classify multiband imagery to distinguish between surfaces like vegetation, soil (pervious), and roads, buildings (impervious) using ArcGIS Pro. This skill is valuable for urban planning, environmental assessment, and stormwater management.

Prerequisites:

Before starting, ensure you have ArcGIS Pro installed on your system and access to multiband imagery of your area of interest.

Step 1: Extracting Bands

Begin by loading your multiband imagery into ArcGIS Pro. Navigate to the Imagery tab and click on Raster Functions. Choose Extract Bands and select the desired bands, such as Near Infrared, Red, and Blue, which can emphasize urban features. Create a new layer with the extracted bands.

Step 2: Configuring Classification Wizard

Select the newly created layer with the extracted bands. Access the Classification Wizard in the Image Classification group under the Imagery tab. Set the Classification Method to Supervised and Classification Type to Object-based. Choose the default schema and specify the output location for the classification results.

Step 3: Segmenting the Image

Adjust segmentation parameters, such as Spectral detail and Spatial detail, to group pixels with similar characteristics into segments. Preview the segmentation results and refine them as needed. Zoom in to inspect the segmentation results in detail.

Step 4: Creating Training Samples

Modify the classification schema to include impervious and pervious classes. Create training samples by drawing polygons on the map for each land-use type. Group similar samples into classes to represent different surface types.

Step 5: Classifying the Image

Train the Support Vector Machine classifier using the defined training samples. Review the classification preview to ensure accuracy. Run the classification process to generate a classified raster dataset. Merge subclasses into parent classes, such as impervious and pervious, to simplify the classification results.

Step 6: Reclassifying Errors

Review the classified raster dataset for any misclassifications. Use the Reclassify within a region tool to correct specific errors, such as misclassified areas. Draw polygons around misclassified areas and reclassify them as needed.

Step 7: Finalizing Classification

Save the final classified raster dataset with appropriate naming conventions. Also, save your ArcGIS Pro project to preserve the classification workflow and results for future reference.

Conclusion:

By following this tutorial, you've successfully classified multiband imagery to identify pervious and impervious surfaces using ArcGIS Pro. This classification can be utilized for various applications such as urban planning, environmental monitoring, and stormwater management. Experiment with different parameters and refine your classification process to achieve accurate results tailored to your specific needs.