ComfyUI Node
Imgutils OCR
Extract text from images using PaddleOCR. Returns recognized text strings and text with confidence scores.
Imgutils OCR
- image
- detections
- json
◄heat_threshold0.30►
◄box_threshold0.70►
Categoryimgutils/ocr
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | Input image for text extraction. | |
| heat_threshold | FLOAT | 0.300–1 | Heat map threshold for text region detection. Lower = more regions. |
| box_threshold | FLOAT | 0.700–1 | Box threshold for text detection. Lower = more candidate boxes. |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| detections | STRING | — |
| json | STRING | — |