OCR Area working NOT as expected #968

New issue

Closed

opened 2026-02-20 14:09:09 -05:00 by deekerman · 1 comment

deekerman commented

2026-02-20 14:09:09 -05:00

Owner

Originally created by @ArifJanMC on GitHub (Dec 1, 2024).

Originally assigned to: @mdevaev on GitHub.

Describe the bug
When using the "Select area" option, the upper boundary of the selected area is detected correctly. However, the bottom boundary is not; instead of limiting to the selected region, the OCR sends the entire portion of the page below the upper boundary for processing.
Is it a known issue? Maybe you can give me a hint on how to fix this?
Tried Chrome & Edge browsers

To Reproduce
Steps to reproduce the behavior, like:

Go to 'Text'
Click on 'Select Area'
Select Area and press "Enter"
Wait until the OCR does its job
Verify that upper border works, while lower does not.

Expected behavior
Area selected on the screen should be the only area which should be given to the OCR lib

Screenshots
Video is better demo.
https://github.com/user-attachments/assets/cf938edb-3f13-4da6-968e-96e5e79228ab

Desktop (please complete the following information):

OS: ANY
Browser ANY
Version ANY
VNC client (if used) -

PiKVM info:

RPi 4
PiKVM platform V3
Video capture type - CSI bridge
KVMD version: kvmd 4.23-1
kvmd-fan 0.31-1
kvmd-platform-v3-hdmi-rpi4 4.23-1
kvmd-webterm 0.50-1
uStreamer version: ustreamer 6.17-1
Linux kernel: Linux pikvm 6.6.45-8-rpi #1 SMP Wed Oct 2 19:56:38 UTC 2024 armv7l GNU/Linux

Additional context
The bug was reproduced and verified by Arch1mede [support].
https://discord.com/channels/580094191938437144/896216359963332628/1311213717912027180

Originally created by @ArifJanMC on GitHub (Dec 1, 2024). Originally assigned to: @mdevaev on GitHub. **Describe the bug** When using the "Select area" option, the upper boundary of the selected area is detected correctly. However, the bottom boundary is not; instead of limiting to the selected region, the OCR sends the entire portion of the page below the upper boundary for processing. Is it a known issue? Maybe you can give me a hint on how to fix this? Tried Chrome & Edge browsers **To Reproduce** Steps to reproduce the behavior, like: 1. Go to 'Text' 2. Click on 'Select Area' 3. Select Area and press "Enter" 4. Wait until the OCR does its job 5. Verify that upper border works, while lower does not. **Expected behavior** Area selected on the screen should be the only area which should be given to the OCR lib **Screenshots** Video is better demo. https://github.com/user-attachments/assets/cf938edb-3f13-4da6-968e-96e5e79228ab **Desktop (please complete the following information):** - OS: ANY - Browser ANY - Version ANY - VNC client (if used) - **PiKVM info:** - RPi 4 - PiKVM platform V3 - Video capture type - CSI bridge - KVMD version: kvmd 4.23-1 kvmd-fan 0.31-1 kvmd-platform-v3-hdmi-rpi4 4.23-1 kvmd-webterm 0.50-1 - uStreamer version: ustreamer 6.17-1 - Linux kernel: Linux pikvm 6.6.45-8-rpi #1 SMP Wed Oct 2 19:56:38 UTC 2024 armv7l GNU/Linux **Additional context** The bug was reproduced and verified by Arch1mede [support]. https://discord.com/channels/580094191938437144/896216359963332628/1311213717912027180