Today, one of the most compelling types of evidence is video. As video evidence becomes more and more common, MD-VIDEO AI is an essential tool for organizations that handle huge volumes of video footage but have limited resources to accurately review the massive amount of data buried in it.
Here we introduce MD-VIDEO AI features that can perform many tasks for you, more accurately and efficiently. The current version of MD-VIDEO AI is highly focused on the detection of various types of objects and colors that are related to the crime scene, this object detection feature is accelerated with a multi-core GPU. And with those labels classified scenes, users can review the summarized results. Also, check out our How-to video on the MD-VIDEO AI and find out how you can be benefited!
How to Run AI Analysis
Check the video to be analyzed and click AI Analysis in the menu.
AI Analysis Settings
Before starting AI analysis, the user can set the options. Select file to be analyzed, preview the selected file and set the object detection type confidence threshold and the frame differential percentage. Also, the date range of each file can be selected by dragging or typing directly.
Display of AI Detected Object
Analysis result consists of File List, Gallery Filter, Event Gallery, and Frame View. Users can directly select the same object and group them and the label can be changed when the label is different from the user’s intention. And if the user is already aware of the color looking for, can set filter only frames that contain that specific color. Filter by selecting objects can be done directly in the Frame View by mouse drawing on the object.
Review on AI Analysis Results
Analysis results are organized into Scene List, Scene Gallery, and Detailed view. Users can remove unnecessary frames from the scene. And the image enhancement feature is supported for the image of frames included in the scene.
Scenes can be merged to create a new scene and unneeded scenes can be removed by pressing Remove. Users can create a movie that contains only scenes. Especially when the video to be analyzed has a long playback time, this function can generate a shortened video containing only meaningful objects.
To learn more about MD-VIDEO AI features, go check our Youtube channel!