fcernafukuzaki
commited on
Commit
•
e806366
1
Parent(s):
6331b5e
Código para YOLOv9
Browse files
app.py
CHANGED
@@ -20,102 +20,114 @@ Este Notebook se acelera opcionalmente con un entorno de ejecución de GPU
|
|
20 |
import os
|
21 |
import re
|
22 |
import json
|
|
|
23 |
import pandas as pd
|
24 |
import gradio as gr
|
25 |
import torch
|
26 |
from PIL import Image
|
|
|
27 |
|
28 |
# Images
|
29 |
-
torch.hub.download_url_to_file('https://
|
30 |
torch.hub.download_url_to_file('https://i.pinimg.com/originals/c2/ce/e0/c2cee05624d5477ffcf2d34ca77b47d1.jpg', 'ejemplo2.jpg')
|
31 |
|
32 |
# Model
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
def
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
|
77 |
-
|
78 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
79 |
|
80 |
def yolo(size, iou, conf, im):
|
81 |
'''Wrapper fn for gradio'''
|
82 |
g = (int(size) / max(im.size)) # gain
|
83 |
im = im.resize((int(x * g) for x in im.size), Image.LANCZOS) # resize with antialiasing
|
84 |
|
|
|
85 |
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
|
|
94 |
|
95 |
-
|
96 |
-
lista = listJSON(results_detail)
|
97 |
-
lista2 = arrayLista(results_detail)
|
98 |
-
return Image.fromarray(results2.ims[0]), lista2, lista
|
99 |
|
100 |
#------------ Interface-------------
|
101 |
|
|
|
|
|
|
|
|
|
102 |
|
103 |
-
|
104 |
-
|
105 |
-
|
106 |
-
in3 = gr.inputs.Slider(minimum=0, maximum=1, step=0.05, default=0.50, label='Umbral o threshold')
|
107 |
-
in4 = gr.inputs.Image(type='pil', label="Original Image")
|
108 |
-
|
109 |
-
out2 = gr.outputs.Image(type="pil", label="YOLOv5")
|
110 |
-
out3 = gr.outputs.Dataframe(label="Cantidad_especie", headers=['Cantidad','Especie'], type="pandas")
|
111 |
-
out4 = gr.outputs.JSON(label="JSON")
|
112 |
#-------------- Text-----
|
113 |
title = 'Trampas Barceló'
|
114 |
-
description =
|
115 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
116 |
article ="<p style='text-align: center'><a href='https://docs.google.com/presentation/d/1T5CdcLSzgRe8cQpoi_sPB4U170551NGOrZNykcJD0xU/edit?usp=sharing' target='_blank'>Para mas info, clik para ir al white paper</a></p><p style='text-align: center'><a href='https://drive.google.com/drive/folders/1owACN3HGIMo4zm2GQ_jf-OhGNeBVRS7l?usp=sharing ' target='_blank'>Google Colab Demo</a></p><p style='text-align: center'><a href='https://github.com/Municipalidad-de-Vicente-Lopez/Trampa_Barcelo' target='_blank'>Repo Github</a></p></center></p>"
|
117 |
|
118 |
-
examples = [['640',0.
|
119 |
|
120 |
iface = gr.Interface(yolo,
|
121 |
inputs=[in1, in2, in3, in4],
|
@@ -125,11 +137,11 @@ iface = gr.Interface(yolo,
|
|
125 |
examples=examples,
|
126 |
analytics_enabled=False,
|
127 |
allow_flagging="manual",
|
128 |
-
flagging_options=["Correcto", "Incorrecto", "Casi correcto", "Error", "Otro"]
|
129 |
-
#flagging_callback=hf_writer
|
130 |
)
|
131 |
-
|
132 |
-
iface.
|
|
|
133 |
|
134 |
"""For YOLOv5 PyTorch Hub inference with **PIL**, **OpenCV**, **Numpy** or **PyTorch** inputs please see the full [YOLOv5 PyTorch Hub Tutorial](https://github.com/ultralytics/yolov5/issues/36).
|
135 |
## Citation
|
|
|
20 |
import os
|
21 |
import re
|
22 |
import json
|
23 |
+
import numpy as np
|
24 |
import pandas as pd
|
25 |
import gradio as gr
|
26 |
import torch
|
27 |
from PIL import Image
|
28 |
+
from ultralytics import YOLO
|
29 |
|
30 |
# Images
|
31 |
+
torch.hub.download_url_to_file('https://i.pinimg.com/originals/7f/5e/96/7f5e9657c08aae4bcd8bc8b0dcff720e.jpg', 'ejemplo1.jpg')
|
32 |
torch.hub.download_url_to_file('https://i.pinimg.com/originals/c2/ce/e0/c2cee05624d5477ffcf2d34ca77b47d1.jpg', 'ejemplo2.jpg')
|
33 |
|
34 |
# Model
|
35 |
+
class YOLODetect():
|
36 |
+
def __init__(self, modelo):
|
37 |
+
self.modelo = modelo
|
38 |
+
|
39 |
+
def predecir(self, url):
|
40 |
+
# conf float 0.25 umbral de confianza del objeto para la detección
|
41 |
+
# iou float 0.7 umbral de intersección sobre unión (IoU) para NMS
|
42 |
+
self.source = url
|
43 |
+
self.results = self.modelo.predict(source=self.source, save=True, imgsz=640, conf=0.5, iou=0.40)
|
44 |
+
return self.results
|
45 |
+
|
46 |
+
def show(self):
|
47 |
+
results = self.results[0]
|
48 |
+
render = None #render_result(model=self.modelo, image=self.source, result=self.results[0])
|
49 |
+
render = Image.open(f"runs/detect/predict/{results.path}")
|
50 |
+
return render
|
51 |
+
|
52 |
+
def to_json(self):
|
53 |
+
results = self.results[0]
|
54 |
+
img_size = results.orig_shape
|
55 |
+
img_name = results.path
|
56 |
+
array_numpy = results.boxes.cls.cpu().numpy().astype(np.int32)
|
57 |
+
|
58 |
+
# Definir las clases y sus nombres correspondientes
|
59 |
+
clases = {
|
60 |
+
0: "Aedes",
|
61 |
+
1: "Mosquitos",
|
62 |
+
2: "Moscas"
|
63 |
+
}
|
64 |
+
|
65 |
+
# Contabilizar las clases
|
66 |
+
conteo_clases = np.bincount(array_numpy)
|
67 |
+
|
68 |
+
self.json_result = [{'Especie': clases[i], 'Cantidad': str(conteo_clases[i]) if i < len(conteo_clases) else str(0)} for i in range(len(clases))]
|
69 |
+
# Crear un diccionario con los elementos necesarios
|
70 |
+
result_dict = {
|
71 |
+
"image": str(img_name),
|
72 |
+
"size": str(img_size),
|
73 |
+
"detail": self.json_result
|
74 |
+
}
|
75 |
+
|
76 |
+
# Convertir el diccionario a una cadena JSON
|
77 |
+
result_dict = json.dumps(result_dict)
|
78 |
+
|
79 |
+
# Convertir la cadena JSON a un objeto Python (diccionario)
|
80 |
+
result_dict = json.loads(result_dict)
|
81 |
+
|
82 |
+
return result_dict
|
83 |
+
|
84 |
+
def to_dataframe(self):
|
85 |
+
return pd.DataFrame(self.json_result)
|
86 |
+
|
87 |
+
modelo_yolo = YOLO('best.pt')
|
88 |
|
89 |
def yolo(size, iou, conf, im):
|
90 |
'''Wrapper fn for gradio'''
|
91 |
g = (int(size) / max(im.size)) # gain
|
92 |
im = im.resize((int(x * g) for x in im.size), Image.LANCZOS) # resize with antialiasing
|
93 |
|
94 |
+
print(type(im))
|
95 |
|
96 |
+
source = im#Image.open(im)
|
97 |
+
model = YOLODetect(modelo_yolo)
|
98 |
+
results = model.predecir(source)
|
99 |
|
100 |
+
result_json = model.to_json()
|
101 |
+
print(result_json)
|
102 |
+
result_df = model.to_dataframe()
|
103 |
+
print(result_df)
|
104 |
+
result_img = model.show()
|
105 |
|
106 |
+
return result_img, result_df, result_json
|
|
|
|
|
|
|
107 |
|
108 |
#------------ Interface-------------
|
109 |
|
110 |
+
in1 = gr.Radio(['640', '1280'], label="Tamaño de la imagen", type='value')
|
111 |
+
in2 = gr.Slider(minimum=0, maximum=1, step=0.05, label='NMS IoU threshold')
|
112 |
+
in3 = gr.Slider(minimum=0, maximum=1, step=0.05, label='Umbral o threshold')
|
113 |
+
in4 = gr.Image(type='pil', label="Original Image")
|
114 |
|
115 |
+
out2 = gr.Image(type="pil", label="YOLOv5")
|
116 |
+
out3 = gr.Dataframe(label="Cantidad_especie", headers=['Cantidad','Especie'], type="pandas")
|
117 |
+
out4 = gr.JSON(label="JSON")
|
|
|
|
|
|
|
|
|
|
|
|
|
118 |
#-------------- Text-----
|
119 |
title = 'Trampas Barceló'
|
120 |
+
description = """
|
121 |
+
<p>
|
122 |
+
<center>
|
123 |
+
Sistemas de Desarrollado por Subsecretaría de Modernización del Municipio de Vicente López. Advertencia solo usar fotos provenientes de las trampas Barceló, no de celular o foto de internet.
|
124 |
+
<img src="https://www.vicentelopez.gov.ar/assets/images/logo-mvl.png" alt="logo" width="250"/>
|
125 |
+
</center>
|
126 |
+
</p>
|
127 |
+
"""
|
128 |
article ="<p style='text-align: center'><a href='https://docs.google.com/presentation/d/1T5CdcLSzgRe8cQpoi_sPB4U170551NGOrZNykcJD0xU/edit?usp=sharing' target='_blank'>Para mas info, clik para ir al white paper</a></p><p style='text-align: center'><a href='https://drive.google.com/drive/folders/1owACN3HGIMo4zm2GQ_jf-OhGNeBVRS7l?usp=sharing ' target='_blank'>Google Colab Demo</a></p><p style='text-align: center'><a href='https://github.com/Municipalidad-de-Vicente-Lopez/Trampa_Barcelo' target='_blank'>Repo Github</a></p></center></p>"
|
129 |
|
130 |
+
examples = [['640',0.45, 0.75,'ejemplo1.jpg'], ['640',0.45, 0.75,'ejemplo2.jpg']]
|
131 |
|
132 |
iface = gr.Interface(yolo,
|
133 |
inputs=[in1, in2, in3, in4],
|
|
|
137 |
examples=examples,
|
138 |
analytics_enabled=False,
|
139 |
allow_flagging="manual",
|
140 |
+
flagging_options=["Correcto", "Incorrecto", "Casi correcto", "Error", "Otro"]
|
|
|
141 |
)
|
142 |
+
|
143 |
+
iface.queue()
|
144 |
+
iface.launch(debug=True)
|
145 |
|
146 |
"""For YOLOv5 PyTorch Hub inference with **PIL**, **OpenCV**, **Numpy** or **PyTorch** inputs please see the full [YOLOv5 PyTorch Hub Tutorial](https://github.com/ultralytics/yolov5/issues/36).
|
147 |
## Citation
|