How to Perform OCR using VIDIZMO Indexer
OCR (Optical Character Recognition) in VIDIZMO allows you to extract text from images, videos, and documents, enhancing their searchability within your portal. Utilizing the VIDIZMO Vision Indexer, you can process your content for OCR detection, extract any detected text, and convert it into searchable metadata within the portal. Additionally, VIDIZMO provides the ability to perform redactions on these detections. Using VIDIZMO's Studio Space, you can selectively conceal sensitive information, such as personally identifiable information (PII), financial data, or other confidential content within your documents, images, and videos. These redacted files are functional and accessible for business use, allowing users to view, search, and interact with them.
Prerequisites
- Ensure you belong to a group with feature permissions for OCR Detection. To configure the VIDIZMO Vision Indexer, you also need the App Management feature enabled.
- For OCR-based redaction using Studio Space, you need to be part of a group with Redaction or Bulk Redaction features enabled.
Configure VIDIZMO Vision Indexer for OCR
- Open the navigation menu via the button on the top left.
- Select the Admin tab.
- Select Portal Settings.
- In Portal Settings, select the Apps tab.
- Select Content Processing to view all the processing applications.
- Select the settings button (gear icon) on the VIDIZMO Vision Indexer to begin its configuration.
- In the VIDIZMO Vision Indexer's settings:
- Select the Media or Evidence Type to perform OCR on.
- Select OCR from the Detection Types dropdown menu. You can also add additional detection types along with OCR.
- Set the Tracking Frame Count (3–30, video only) to define how many consecutive frames an object must appear in.
- Set the Frame Forward Count (1–15) to control how many frames are skipped after processing a frame.
- Set the OCR Frame Forward Count (1–30) to control how many frames are skipped after OCR processing.
- Choose the OCR Language that the Vision Indexer will detect.
- To automatically generate OCR for your content as it is uploaded to your Portal, set Automatic Processing to on.
- Select Save Changes to save your configurations.
- Enable the VIDIZMO Vision Indexer using the toggle button.
Searching OCR-Generated Content Using Portal Search
- Type the keyword you want to search for and select the search icon. If the keyword is present in an OCR detection, the matching items are displayed.
- Select the desired occurrence of your searched keyword that is displayed against the relevant media. You are redirected to the instance where that specific character appeared.
Searching OCR-Generated Content within Playback
- From the media playback Insights tab, search for your desired keyword.
- From the search results, select any OCR detection to be redirected to the exact page (for documents) or the specific timestamp (for videos) where your searched term appears.
Redaction of OCR Objects
Using Studio Space, VIDIZMO enables users to perform OCR-based redactions for videos, images, and documents. Studio Space provides tools for precise and selective redaction of detected objects, including text identified through Optical Character Recognition (OCR). To learn more about the redaction capabilities of Studio Space, see Redaction using Studio Space: A Comprehensive Walkthrough.
To redact OCR objects in a video (or image):
- Select Redact from the file's overflow menu to open it in Studio Space.
- If OCR has not been generated for your content, select Auto Detect from the header.
- In the Start Scan window, add OCR to the Detection Types.
- Select Start to process your file for OCR.
- Once processing is complete, the OCR objects are generated and displayed in the left panel. Select the objects you want to redact via their corresponding checkboxes.
- After making your selection, select Redact.
- To apply redactions to the same file, select Redact. To create a copy of your file with the redactions applied, select Redact as New.
- After completing the redaction processing, the file appears in your Portal library with a Redacted tag. Select the file to play it back.
- The selected OCR objects are redacted.
To perform document redaction using OCR, see How to Perform Document Redaction in VIDIZMO.