OCI Stored Video Analysis
OCI Video Analysis is a part of the OCI Vision service, designed to process stored videos and identify labels, objects, texts, and faces within each frame. It can analyze the entire video or every frame in the video using pre-trained or custom models. The feature provides information about the detected labels, objects, texts, and faces, and provides the time at which they’re detected. With a timeline bar, you can directly look for a label or object and navigate to the exact timestamp in the video where a particular label or object is found.
To use this feature, you’ll need to upload your video to an OCI Bucket. Here is an example of a video stored in a bucket called OCI-Vision-Video-Demos.
You might need to allow Pre-Authenticated Requests for this bucket. If you don’t do this, you will be prompted by the tool to allow this.
Next, go to the OCI Vision page. You’ll find the Video Analysis tool at the bottom of the menu on the left-hand side of the page.
You can check out the demo videos, or load your own video from the Local File system of you computer, or use a file from your OCI Storage. If you select a video from the Local File system, the video will be loaded in the object storage before it is processed.
For this example, I’m going to use the video I uploaded earlier called Trinity-Student.mp4. Copy the link to this file from the file information in object storage.
On the Video Analysis page, select Object Storage and paste the link to the file into the URL field. Then click Analyze button. It is at this point that you might get asked to Generate a PAR URL. Do this and then proceed.
While the video is being parsed, it will appear on the webpage and will start playing. When the video has been Analyzed the details will be displayed below the video. The Analysis will consist of Label Detection, Object Detection, Text Dection and Face Detection.
By clicking on each of these, you’ll see what has been detected, and by clicking on each of these, you’ll be able to see where in the video they were detected. For example, where a Chair was detected.
You can also inspect the JSON file containing all the details of various objects detected in the video and the location in the videos they can be found.
This JSON response file is also saved to Object Storage in the same directory, or a sub-directory, where the video is located.
Ora-lytics 
