Inputs
image: Input image for OCR.
01
float32?: Minimum confidence threshold for recognized text.
01
float32?: Minimum score for a region to be detected as text.
01
float32?: Minimum score to apply orientation correction to a text region.
01
float32?: Minimum score for a detected box to be kept.
int32?: Maximum side length for the detection input image.
float32?: Ratio to expand detected text boxes.
12000
int32?: Maximum number of text region candidates to process.
Press and hold for realtime mode.
Outputs
[]
list: OCR results with text, confidence, and bounding boxes.