← Models

CV / PaddleOCR

PaddleOCR v5 Mobile Recognition

reComputer R Series (CM5 + Hailo-8)

Text-line recognition on reComputer R Series with Hailo-8 or Hailo-10H.

54 downloads
Size
7.48 MB
Memory
4GB+
Precision
Hailo HEF / HailoRT

Choose the device you're using, the set up guide and documentation will update accordingly.

Getting Started

Deploy
sudo docker run --rm \
  --name cm5-hailo8-paddle-ocr-v5-mobile-recognition \
  --privileged \
  --net=host \
  -e PYTHONUNBUFFERED=1 \
  --device /dev/hailo0:/dev/hailo0 \
  -v /usr/lib/libhailort.so.4.23.0:/usr/lib/libhailort.so.4.23.0:ro \
  -v /usr/lib/libhailort.so:/usr/lib/libhailort.so:ro \
  ghcr.io/seeed-projects/recomputer-hailo8-cv/paddle_ocr_v5_mobile_recognition:latest \
  python web_detection.py --model_path model/paddle_ocr_v5_mobile_recognition.hef --video_path video/test.mp4

Model Details

reComputer R Series

PaddleOCR v5 Mobile Recognition on reComputer R Series (CM5 + Hailo-8)

PaddleOCR v5 Mobile Recognition decodes one cropped text-line image. It is an independent model and does not locate text regions in a full document; use a text detector first when processing a document page.

Model information

PropertyValue
TaskText recognition
Input48 × 320 × 3 cropped text-line image
OutputCTC-decoded text and confidence
HEFHailo-8 PaddleOCR recognition HEF
RuntimeHailoRT 4.23.x

Host requirements

The Hailo-8 driver, firmware, host libhailort.so, and Python wheel inside the image must use matching HailoRT major/minor versions.

bash
sudo apt update
sudo apt install hailort hailort-pcie-driver python3-hailort
sudo reboot

hailortcli --version
hailortcli fw-control identify
ls -l /dev/hailo0

Run the demo

bash
sudo docker run --rm \
  --name cm5-hailo8-paddle-ocr-v5-mobile-recognition \
  --privileged \
  --net=host \
  -e PYTHONUNBUFFERED=1 \
  --device /dev/hailo0:/dev/hailo0 \
  -v /usr/lib/libhailort.so.4.23.0:/usr/lib/libhailort.so.4.23.0:ro \
  -v /usr/lib/libhailort.so:/usr/lib/libhailort.so:ro \
  ghcr.io/seeed-projects/recomputer-hailo8-cv/paddle_ocr_v5_mobile_recognition:latest \
  python web_detection.py --model_path model/paddle_ocr_v5_mobile_recognition.hef --video_path video/test.mp4

Open http://<PI_IP>:8000 for the preview.

REST API

bash
curl -X POST "http://<PI_IP>:8000/api/models/paddle_ocr_v5_mobile_recognition/predict" \
  -F "file=@text-line.png"

The response includes the decoded text and its CTC confidence.

Relationship to detection

This model expects a single text line. Send full document images to PaddleOCR v5 Mobile Detection first, then crop each returned region before recognition.

Inputs and Outputs

Input: cropped single-line text image or demo video frame. Output: CTC-decoded text, confidence, and annotated MJPEG preview.