Quick Start

This section will take you through AIVT-OV The basic use of the toolkit, including image acquisition, video playback, model inference and other typical tasks, helps you quickly establish the first AI Visual applications.

1. image acquisition and visual processing (Computer Vision)

1.1 Read and save pictures

  1. Open the sample path: Examples can be found inLabVIEW install path\examples\VIRobotics\found in the path. It can also be found in the help of the LabVIEW, as follows: Help>>Find Examples>>Directory Structure>>VIRobotics>>AI Vision>>CV>>Image Reading and Capture

  1. Double-click the example Image Display And Save.vi

  1. Pictures can be displayed and saved after running

4, press Ctrl + E View Block Diagram :

1.2 Capture camera image

  1. Make sure you are connected USB The camera or computer has a built-in camera;

  2. Open the example Capture Video From Camera.vi;

  3. Run the program to obtain real-time pictures

  1. The program block diagram is shown in the following figure:

1.3 Read webcam or video streams

1, open. Capture Video From Camera.vi;

  1. Switch the source control to URL and enter the video stream address, which can be an online URL link or a webcam URL link, such as https://video.wezhan.cn/90e852575ba871f0bffe4531958d0102/59c6bd1b28784e2ba99c0ac28e13a477-88252280e192c3a9e71392ae3f5bec72-sd.mp4

3, can refer Play Video From File.vi Realize screen zoom display

1.4 Read and save video files

  1. Open the example Play Video From File.vi: Play local video files

  2. Open the example Save Video.viSave the camera screen (or video stream) as a video file ;

2. model loading and inference (AI Model Inference)

The following will be YOLOv8 The model is used as an example to demonstrate how to load the model for target detection.

⚠️ Note: Make sure that the model path and image path are in English and avoid containing Chinese and spaces.

2.1 Real-time detection (via camera)

  1. Open Sample Path: Examples are available inLabVIEW install path\examples\VIRobotics\found in the path. It can also be found in the help of the LabVIEW, as follows: Help>>Find Examples>>Directory Structure>>VIRobotics>>AI Vision>>YOLOv8

  2. Double click to open YOLOv8_OpenVINO.vi;

  3. Click Run, default load yolov8 model and call the camera to detect

2.2 Picture Detection (Detecting Picture Sets)

  1. Open Sample YOLOv8_OpenVINO_imgs.vi

  2. The program will read the specified picture folder, click on the right side imgs Control to switch detection pictures

2.3 Replace with custom model

You can replace the default model in the above example with your own trained model:

ProjectDescription
model_fileReplace with your own .onnx or .xml(IR) Model Path
class_names_fileReplace with a category label file (such .txt)
image FolderImage directory path used for image detection

After the settings are completed, click Run to complete the inference of the custom model.

Tips and Precautions

  • All paths must be in English, avoid Chinese/spaces/special characters

  • The default model is yolov8s.onnx, if you replace the model, please ensure that the format is compatible.

  • Please refer Function Module Details Get each VI Detailed parameter description

Technical Support

If you encounter problems during use, please refer to the relevant chapters or contact technical support: