您最多选择25个主题 主题必须以中文或者字母或数字开头,可以包含连字符 (-),并且长度不得超过35个字符

3.2 KiB

Getting Started with SynthDet

This will provide a step by step instructions on creating a new scene using the Perception features to create semantic data and image captures. These steps should work with both options for setup steps of using a existing project or creating a new project. The goal is to have a working scene by the end of these instructions that will provide you with a dataset, rgb images captures, and segmentic data.

If you have not already done the setup steps for the project, click here to start project setup.

Step 1: Create a new scene and camera

  1. Create a new scene File-> New Scene
  2. Save the Scene File-> Save and give it a name, i.e created scene name is PerceptionScene
  3. Select the Main Camera and reset the Position transform to 0
  4. In the Hierarchy window select the main camera
    1. In the inspector panel of the main camera select Add Component
    2. Add a Perception Camera component

Step 2: Create labeled objects

  1. In the Hierarchy window right click -> Go to 3D Object -> Select Cube
    1. Create 3 Cubes
    2. Change the names of the cubes to have 3 seperate names Cube, Box, Crate
    3. Position the Cubes in front of the FOV of the main Camera, example image of the completed scene for reference down below
  2. For each object in the scene that was created, from the inspector panel add the script called Labeling
    1. Click the +
    2. In the text field add the name of the object i.e Crate
  3. In the Project panel right click -> Perception -> Labeling Configuration
  4. Select the Labeling Configuration created in the project panel
    1. Click the +
    2. In the label text field add the same text that the Label script contains on the objects created in the scene (i.e Cube, Box, Crate)
    3. Add a numerical value to the value field
      1. Make sure the labels all have different values, for this example use values of 10,000
  5. Select the Main Camera in the Hierarchy panel
    1. In the Perception Camera script in the Labeling Configuration field add the Labeling Configuration script created in previous step

Step 3: Checking local files

  1. Press play in the editor and allow the scene to run for 10 seconds before ending playmode
  2. In the console log you will see a Shutdown in Progress message that will show a file path to the location of the generated dataset
  3. The file path is the Application Persistent Path + /Defaultcompany/UnityTestFramework/
    1. Example file path on a Windows PC : C:/Users//AppData/LocalLow/DefaultCompany/UnityTestFramework\2e10ec21-9d97-4cee-b5a2-7e95e299afa4\RGB18f61842-ef8d-4b31-acb5-cb1da36fb7b1
  4. In the output path for the Labeling content you can verify the following data is present:
    1. RGB captures
    2. Semantic segmentation images
    3. Logs
    4. JSON Dataset