app.py

# import os
# import yolov5

# # load model
# model = yolov5.load('keremberke/yolov5m-license-plate')

# # set model parameters
# model.conf = 0.5  # NMS confidence threshold
# model.iou = 0.25  # NMS IoU threshold
# model.agnostic = False  # NMS class-agnostic
# model.multi_label = False  # NMS multiple labels per box
# model.max_det = 1000  # maximum number of detections per image

# # set image
# def license_plate_detect(img):
#     # perform inference
#     results = model(img, size=640)
    
#     # inference with test time augmentation
#     results = model(img, augment=True)
    
#     # parse results
#     if len(results.pred):
#         predictions = results.pred[0]
#         boxes = predictions[:, :4] # x1, y1, x2, y2
#         scores = predictions[:, 4]
#         categories = predictions[:, 5]
#         return boxes

# from PIL import Image
# # image = Image.open(img)
# import pytesseract

# def read_license_number(img):
#     boxes = license_plate_detect(img)
#     if boxes:
#         return [pytesseract.image_to_string(
#                     image.crop(bbox.tolist()))
#                for bbox in boxes]

# from transformers import CLIPProcessor, CLIPModel
# vit_model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
# processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")

# def zero_shot_classification(image, labels):
#     inputs = processor(text=labels,
#                        images=image,
#                        return_tensors="pt",
#                        padding=True)
#     outputs = vit_model(**inputs)
#     logits_per_image = outputs.logits_per_image # this is the image-text similarity score
#     return logits_per_image.softmax(dim=1) # we can take the softmax to get the label probabilities

# installed_list = []
# # image = Image.open(requests.get(url, stream=True).raw)
# def check_solarplant_installed_by_license(license_number_list):
#     if len(installed_list):
#         return [license_number in installed_list
#                 for license_number in license_number_list]

# def check_solarplant_installed_by_image(image, output_label=False):
#     zero_shot_class_labels = ["bus with solar panel grids",
#                               "bus without solar panel grids"]
#     probs = zero_shot_classification(image, zero_shot_class_labels)
#     if output_label:
#         return zero_shot_class_labels[probs.argmax().item()]
#     return probs.argmax().item() == 0

# def check_solarplant_broken(image):
#     zero_shot_class_labels = ["white broken solar panel",
#                               "normal black solar panel grids"]
#     probs = zero_shot_classification(image, zero_shot_class_labels)
#     idx = probs.argmax().item()
#     return zero_shot_class_labels[idx].split(" ")[1-idx]

from fastsam import FastSAM, FastSAMPrompt
os.system('wget https://huggingface.co/spaces/An-619/FastSAM/resolve/main/weights/FastSAM.pt')

model = FastSAM('./FastSAM.pt')
DEVICE = 'cpu'
def segment_solar_panel(img):
    # os.system('python Inference.py --model_path FastSAM.pt --img_path bus.jpg  --text_prompt "solar panel grids"')
    img = img.convert("RGB")

    everything_results = model(img, device=DEVICE, retina_masks=True, imgsz=1024, conf=0.4, iou=0.9,)
    prompt_process = FastSAMPrompt(img, everything_results, device=DEVICE)
    
    # everything prompt
    ann = prompt_process.everything_prompt()
    
    # bbox default shape [0,0,0,0] -> [x1,y1,x2,y2]
    ann = prompt_process.box_prompt(bbox=[[200, 200, 300, 300]])
    
    # text prompt
    ann = prompt_process.text_prompt(text='solar panel grids')
    
    # point prompt
    # points default [[0,0]] [[x1,y1],[x2,y2]]
    # point_label default [0] [1,0] 0:background, 1:foreground
    ann = prompt_process.point_prompt(points=[[620, 360]], pointlabel=[1])
    
    prompt_process.plot(annotations=ann,output_path='./bus.jpg',)
    return Image.Open('./bus.jpg')


import gradio as gr

def greet(img):
    lns = read_license_number(img)
    if len(lns):
        seg = segment_solar_panel(img)
        return (seg, '')
        # return (seg,
        #         "車牌： " + '; '.join(lns) + "\n\n" \
        #         + "類型： "+ check_solarplant_installed_by_image(img, True) + "\n\n" \
        #         + "狀態：" + check_solarplant_broken(img))
    return (img, "空地。。。")

iface = gr.Interface(fn=greet, inputs="image", outputs=["image", "text"])
iface.launch()