app.py

import glob

import gradio as gr
import yolov5

model = yolov5.load('model/dango.pt')

#てすと

def inference(gr_input):
    # set model parameters
    model.conf = 0.45  # NMS confidence threshold
    model.iou = 0.45  # NMS IoU threshold
    model.agnostic = False  # NMS class-agnostic
    model.multi_label = False  # NMS multiple labels per box
    model.max_det = 1  # maximum number of detections per image
    results = model(gr_input)
    output_folder = "results"
    if not os.path.exists(output_folder):
       os.makedirs(output_folder)
    # parse results
    predictions = results.pred[0]
    boxes = predictions[:, :4]  # x1, y1, x2, y2
    scores = predictions[:, 4]
    categories = predictions[:, 5]

    # boxesからx1,y1,x2,y2を取り出す
    x1 = boxes[:, 0].tolist()
    y1 = boxes[:, 1].tolist()
    x2 = boxes[:, 2].tolist()
    y2 = boxes[:, 3].tolist()

    # 4つの座標のうち、一つでも入っていなかったら、その画像はスキップ
    if x1 == [] or y1 == [] or x2 == [] or y2 == []:
        return

    x1 = int(x1[0])
    y1 = int(y1[0])
    x2 = int(x2[0])
    y2 = int(y2[0])

    img = Image.open(image)
    img_crop = img.crop((x1, y1, x2, y2))
    img_name = os.path.basename(image)
    # pngで保存
    
    results.save(save_dir='results/')
    img_crop.save(f"{output_folder}/{img_name}", quality=95)
    img_list = [glob.glob("result/*.png")]
    return img_list


with gr.Blocks() as app:
    gr.Markdown('<img id="visitor-badge" alt="visitor badge" src="https://visitor-badge.glitch.me/badge?page_id=alrab222.Cinderella" />')
    gr.Markdown(
                "# <center> ダンゴムシ捕捉\n"
                "## <center> ダンゴムシの腹側からの画像を、機械学習で判別できるモデルです\n"
    )
    inputs = gr.Image()
    output = gr.Gallery(label="結果")
    btn = gr.Button("judge")
    btn.click(inference, inputs, output)
app.launch()