utils.py

import numpy
import sahi.predict
import sahi.utils
import PyPDF4
from pdf2image import convert_from_path
from PIL import Image

TEMP_DIR = "temp"


def sahi_yolov8m_inference(
    image,
    detection_model,
    slice_height=512,
    slice_width=512,
    overlap_height_ratio=0.2,
    overlap_width_ratio=0.2,
    image_size=640,
    #postprocess_type="GREEDYNMM",
    #postprocess_match_metric="IOS",
    postprocess_match_threshold=0.5,
    #postprocess_class_agnostic=False,
):

    # standard inference
    detection_model.image_size = image_size
    prediction_result_1 = sahi.predict.get_prediction(
        image=image, detection_model=detection_model
    )
    visual_result_1 = sahi.utils.cv.visualize_object_predictions(
        image=numpy.array(image),
        object_prediction_list=prediction_result_1.object_prediction_list,
    )
    output_1 = Image.fromarray(visual_result_1["image"])

    # sliced inference
    prediction_result_2 = sahi.predict.get_sliced_prediction(
        image=image,
        detection_model=detection_model,
        slice_height=slice_height,
        slice_width=slice_width,
        overlap_height_ratio=overlap_height_ratio,
        overlap_width_ratio=overlap_width_ratio,
        #postprocess_type=postprocess_type,
        #postprocess_match_metric=postprocess_match_metric,
        postprocess_match_threshold=postprocess_match_threshold,
        #postprocess_class_agnostic=postprocess_class_agnostic,
    )
    visual_result_2 = sahi.utils.cv.visualize_object_predictions(
        image=numpy.array(image),
        object_prediction_list=prediction_result_2.object_prediction_list,
    )

    output_2 = Image.fromarray(visual_result_2["image"])

    return output_1, output_2

def convert_pdf_file(
    path,
    #filename=name,
    dpi=300,
    image_width=4961,
    image_heigth=3508,
    grayscale=True,
):
    with open(path, 'rb') as pdf_file:
        pdf_reader = PyPDF4.PdfFileReader(pdf_file, strict=False)
        first_page = pdf_reader.getPage(0)
        page_size = (first_page.mediaBox.getWidth(), first_page.mediaBox.getHeight())
    if page_size[0] > page_size[1]:
        image = convert_from_path(path, dpi=dpi, size=(image_width,image_heigth), grayscale=grayscale)
    else:
        image = convert_from_path(path, dpi=dpi, size=(image_heigth,image_width), grayscale=grayscale)
    return image
    #image[0].save(f'{path}/{filename}.png', 'PNG')