Text recognition (OCR)

On this page it is possible to choose the method and set up optical character recognition (OCR) features used in DLP-analysis and should be turned on at this page.

Built-in
Specify required languages, separated by commas: aze (Azer), eng (English), kaz (Kazakh), rus (Russian), tur (Turkish), ukr (Ukrainian), uzb (Uzbek).
Attention! Adding each new language to the list significantly slows down the recognition speed!
Maximum number of threads actually determines the speed of parallel processing of multiple requests to the server. If you specify 0, the number will be selected automatically. Specifying a value greater than the physical number of threads of the server CPU does not make sense. It is important to note that each thread at peak can take up to 100 MB of RAM!

Enable faces detection
You can enable this option as an addition to OCR to detect the presence of facial photos in documents or screenshots.
The option must be used simultaneously with adding the corresponding face template (FACE) on the page "DLP - Rules"
Enabling this option will slightly slow down the OCR process!

Attention! After changing the values on this settings page, it is recommended to restart the server service for the new settings to take effect!

v11.1.3476 (build: Jan 12 2026)
Introduction +Software suite structure +Software suite installation +Software suite uninstall +Software suite update -Global settings Database users -Software suite settings General settings description -Server settings Common settings Postponed monitoring Monitoring - Screenshots Monitoring - Webcams Monitoring - Autorecording Monitoring - Printing Monitoring - Shadow copy Monitoring - File hashes Monitoring - Users online Monitoring - Global search Monitoring - Chats-calls Face recognition Text recognition (OCR) Text classification Neural network server LLM-server Azure-integration Webex-integration Reports generator - Parameters Reports generator - Reports (for bosses) Reports generator - Reports (for employees) Reports generator - Saving to folder Reports generator - Sending via FTP Reports generator - Sending by e-mail Reports generator - Sending to website Reports generator - Sending to file sharing Reports generator - Threats Notifications generator - Sending by e-mail Notifications generator - Sending by SMS Notifications generator - Integration with Telegram Notifications generator - 2FA (BOSS) Client protection Events Regular expressions Work schedule syslog Web-interface +Client settings (computer) +Client settings (user) Groups Company structure Work schedules Dossier of employees Sync with AD Risk analyzer Report templates File hashes Tariffs List of users Work with DB SQL-console Journal +Other +FAQ +Technical support	Text recognition (OCR) On this page it is possible to choose the method and set up optical character recognition (OCR) features used in DLP-analysis and should be turned on at this page. Built-in Specify required languages, separated by commas: aze (Azer), eng (English), kaz (Kazakh), rus (Russian), tur (Turkish), ukr (Ukrainian), uzb (Uzbek). Attention! Adding each new language to the list significantly slows down the recognition speed! Maximum number of threads actually determines the speed of parallel processing of multiple requests to the server. If you specify 0, the number will be selected automatically. Specifying a value greater than the physical number of threads of the server CPU does not make sense. It is important to note that each thread at peak can take up to 100 MB of RAM! Enable faces detection You can enable this option as an addition to OCR to detect the presence of facial photos in documents or screenshots. The option must be used simultaneously with adding the corresponding face template (FACE) on the page "DLP - Rules" Enabling this option will slightly slow down the OCR process! Attention! After changing the values on this settings page, it is recommended to restart the server service for the new settings to take effect!
© KICKIDLER DLP