Spaces:
Running
Running
File size: 2,597 Bytes
2fdf343 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 |
import os
import gradio as gr
from imgutils.detect import detect_person, detect_halfbody, detect_heads, detection_visualize
def _split_image(image, head_scale: float):
retval = []
all_detects = []
for i, (px, _, score) in enumerate(detect_person(image), start=1):
person_image = image.crop(px)
person_label = f'Person #{i} ({score * 100.0:.1f}%)'
retval.append((person_image, person_label))
all_detects.append((px, 'person', score))
px0, py0, _, _ = px
half_detects = detect_halfbody(person_image)
if half_detects:
halfbody_image = person_image.crop(half_detects[0][0])
halfbody_label = f'Person #{i} - Half Body'
retval.append((halfbody_image, halfbody_label))
bx0, by0, bx1, by1 = half_detects[0][0]
all_detects.append(((bx0 + px0, by0 + py0, bx1 + px0, by1 + py0), 'halfbody', half_detects[0][2]))
head_detects = detect_heads(person_image)
if head_detects:
(hx0, hy0, hx1, hy1), _, head_score = head_detects[0]
cx, cy = (hx0 + hx1) / 2, (hy0 + hy1) / 2
width, height = hx1 - hx0, hy1 - hy0
width = height = max(width, height) * head_scale
x0, y0 = int(max(cx - width / 2, 0)), int(max(cy - height / 2, 0))
x1, y1 = int(min(cx + width / 2, person_image.width)), int(min(cy + height / 2, person_image.height))
head_image = person_image.crop((x0, y0, x1, y1))
head_label = f'Person #{i} - Head'
retval.append((head_image, head_label))
all_detects.append(((x0 + px0, y0 + py0, x1 + px0, y1 + py0), 'head', head_score))
return detection_visualize(image, all_detects), retval
if __name__ == '__main__':
with gr.Blocks() as demo:
with gr.Row():
with gr.Column():
gr_input = gr.Image(type='pil', label='Original Image')
gr_head_scale = gr.Slider(0.8, 2.5, 1.5, label='Head Scale')
gr_button = gr.Button(value='Crop', variant='primary')
with gr.Column():
with gr.Tabs():
with gr.Tab('Detected'):
gr_detected = gr.Image(type='pil', label='Detection')
with gr.Tab('Cropped'):
gr_gallery = gr.Gallery(label='Cropped Images')
gr_button.click(
_split_image,
inputs=[gr_input, gr_head_scale],
outputs=[gr_detected, gr_gallery],
)
demo.queue(os.cpu_count()).launch()
|