Description #

Multilingual OCR recognition component based on EasyOCR, supporting text detection and recognition in mixed Chinese and English scenarios.

Key features #

• Supports recognition of more than 100 languages (requires configuration of the corresponding language model)
• Input format compatibility:
  • TOS URL
  • Base64 encoding
  • Binary stream
  • Numpy array
• Performance optimization:
  • GPU-accelerated inference
  • Model quantization (enabled by default)
  • Batch processing optimization

Multilingual support #

• Simplified Chinese (ch_sim)
• English (en)
• Japanese (ja)
• Korean (ko)
• French (fr)
• German (de)

For the complete language list, refer to the official documentation: https://www.jaided.ai/easyocr/

Operator parameters #

Input #

Output #

An array containing OCR recognition results, with each element being a string

Parameters #

Parameters without a default value are required.

Examples #

The following code demonstrates how to use daft to recognize text in images.

from __future__ import annotations

import os

import daft
from daft import col
from daft.las.functions.image.image_easyocr import ImageEasyOcr
from daft.las.functions.udf import las_udf

if __name__ == "__main__":
    if os.getenv("DAFT_RUNNER", "native") == "ray":
        import logging

        import ray

        def configure_logging():
            logging.basicConfig(
                level=logging.INFO,
                format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
                datefmt="%Y-%m-%d %H:%M:%S.%s".format(),
            )
            logging.getLogger("tracing.span").setLevel(logging.WARNING)
            logging.getLogger("daft_io.stats").setLevel(logging.WARNING)
            logging.getLogger("DaftStatisticsManager").setLevel(logging.WARNING)
            logging.getLogger("DaftFlotillaScheduler").setLevel(logging.WARNING)
            logging.getLogger("DaftFlotillaDispatcher").setLevel(logging.WARNING)

        ray.init(dashboard_host="0.0.0.0", runtime_env={"worker_process_setup_hook": configure_logging})
        daft.set_runner_ray()
    daft.set_execution_config(actor_udf_ready_timeout=600)
    daft.set_execution_config(min_cpu_per_task=0)

    tos_dir_url = os.getenv("TOS_DIR_URL", "las-cn-beijing-public-online.tos-cn-beijing.volces.com")
    samples = {
        "image": [
            f"https://{tos_dir_url}/public/shared_image_dataset/.jpeg"
        ]
    }

    image_src_type = "image_url"
    model_path = os.getenv("MODEL_PATH", "/opt/las/models")
    model_name = "EasyOCR"
    quantize = True
    lang_list = ["en", "ch_sim"]
    batch_size = 16
    num_gpus = 1

    ds = daft.from_pydict(samples)
    ds = ds.with_column(
        "ocr_result",
        las_udf(
            ImageEasyOcr,
            construct_args={
                "image_src_type": image_src_type,
                "model_path": model_path,
                "model_name": model_name,
                "quantize": quantize,
                "lang_list": lang_list,
                "batch_size": batch_size,
            },
            num_gpus=num_gpus,
            batch_size=1,
        )(col("image")),
    )
    ds.show()
    df = ds.to_pandas()

    # ╭────────────────────────────────┬───────────────────╮
    # │ image                          ┆ ocr_result        │
    # │ ---                            ┆ ---               │
    # │ Utf8                           ┆ Utf8              │
    # ╞════════════════════════════════╪═══════════════════╡
    # │ tos://las-cn-beijing-public-o… ┆ No matter the outcome           │
    # │                                ┆ I am already very grateful for having met…    │
    # ╰────────────────────────────────┴───────────────────╯