How to Create Professional Quality 3D Models from Photos with Agisoft PhotoScan Professional 1.3.0
Agisoft PhotoScan Professional 1.3.0 is an advanced image-based 3D modeling solution that allows you to create high-quality 3D content from still images. Whether you want to reconstruct archaeological sites, artifacts, buildings, interiors, people, or dynamic scenes, PhotoScan Professional can help you achieve realistic and accurate results.
In this article, we will show you how to use PhotoScan Professional 1.3.0 to create a 3D model from a set of photos taken from different angles. We will also introduce some of the features and benefits of this software, such as automatic image alignment, dense point cloud generation and editing, digital elevation model (DEM) creation and editing, georeferenced orthomosaic generation, 4D modeling for dynamic scenes, panorama stitching, and powerlines detection.
Step 1: Import Photos
The first step is to import the photos that you want to use for your 3D model. You can use any type of images, such as aerial (nadir, oblique), close-range, or satellite photos. You can also use scanned images with fiducial marks or multi-camera rig data for creative projects.
To import photos, go to Workflow > Add Photos and select the folder or files that contain your images. You can also drag and drop the images into the Workspace pane. PhotoScan Professional will automatically detect the camera parameters and EXIF metadata of your images and display them in the Photos pane.
Step 2: Align Photos
The next step is to align the photos and generate a sparse point cloud that represents the approximate geometry of the scene. PhotoScan Professional uses a multi-view 3D reconstruction technology that operates with arbitrary images and is efficient in both controlled and uncontrolled conditions.
To align photos, go to Workflow > Align Photos and adjust the settings according to your needs. You can choose the accuracy level, the generic or adaptive preselection mode, the key point limit, and the tie point limit. You can also enable or disable mask generation, reference preselection, or camera calibration estimation.
After aligning photos, you can view the sparse point cloud and the camera positions in the Model view. You can also check the alignment quality and accuracy in the Reference pane.
Step 3: Build Dense Point Cloud
The third step is to build a dense point cloud that represents the detailed geometry of the scene. PhotoScan Professional uses a depth map-based approach that produces high-resolution point clouds with millions of points.
To build a dense point cloud, go to Workflow > Build Dense Cloud and adjust the settings according to your needs. You can choose the quality level, the depth filtering mode, and the point colors source. You can also enable or disable GPU acceleration or network processing.
After building a dense point cloud, you can view it in the Model view and edit it using various tools in the Tools menu. You can also classify points into different categories such as ground, building, vegetation, etc., using the Classify Points tool.
Step 4: Build Mesh
The fourth step is to build a mesh that represents the surface of the scene. PhotoScan Professional uses a surface reconstruction algorithm that creates a polygonal mesh from the dense point cloud.
To build a mesh, go to Workflow > Build Mesh and adjust the settings according to your needs. You can choose the surface type (arbitrary or height field), the source data (dense cloud or depth maps), and the face count (low, medium, high, custom).
After building a mesh, you can view it in the Model view and edit it using various tools in the Tools menu. You can also export it in various formats such as OBJ, PLY, STL, etc., using the Export > Export Model command.
Step 5: Build Texture
The fifth step is to build a texture that represents the color of the scene. PhotoScan Professional uses a photorealistic texturing algorithm that creates a texture map from the original images.