Skip to content

Quick Start

1. Installation

pip install rapidocr onnxruntime

2. Usage

rapidocr -img "https://www.modelscope.cn/models/RapidAI/RapidOCR/resolve/master/resources/test_files/ch_en_num.jpg" --vis_res

from rapidocr import RapidOCR

engine = RapidOCR()
img_url = "https://www.modelscope.cn/models/RapidAI/RapidOCR/resolve/master/resources/test_files/ch_en_num.jpg"
result = engine(img_url)

print(result.txts)
print(result.scores)
result.vis("vis_result.jpg")

Note

result.vis() downloads the font required for visualization on first use. In an offline environment, omit this line or configure font_path in advance.

3. View the visualization result

4. Inspect the result

result is a RapidOCROutput data class. The most commonly used fields are:

Field Type Description
boxes np.ndarray Four-point coordinates for each text line, shaped (N, 4, 2)
txts Tuple[str] Recognized text, in the same order as boxes
scores Tuple[float] Confidence score for each text line
elapse float Total inference time in seconds 
print(result.boxes.shape)
print(result.txts)
print(result.scores)

See the usage guide for the complete output structure and stage-specific results.

Support for other programming languages

Comments