Lab03 - Histograma e equalização

Objetivos da aula:

• Histograma e equalização de histograma
• Seguimentação com auxilio do histograma
• Webcam opencv

Histograma¶

Na ultima aula, tentamos fazer a segmentação de um objeto da imagem pelo metodo força bruta #GoHorse, pode ser que funcione mas não é a forma mais intessnte de ser feita. Um histograma pode nos ajudar, ele plota em um gráfico de frequência as componentes de cores (r,g,b ou gray) da imagem.

In [ ]:

%matplotlib inline
import cv2
from matplotlib import pyplot as plt
import numpy as np

print ("OpenCV Version : %s " % cv2.__version__)

%matplotlib inline import cv2 from matplotlib import pyplot as plt import numpy as np print ("OpenCV Version : %s " % cv2.__version__)

OpenCV Version : 4.6.0 

In [ ]:

!wget "https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/PDI/lab03/bola.png" /content
!wget "https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/PDI/lab03/bolinha.png" /content
!wget "https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/PDI/lab03/fuca.png" /content

!wget "https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/PDI/lab03/bola.png" /content !wget "https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/PDI/lab03/bolinha.png" /content !wget "https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/PDI/lab03/fuca.png" /content

--2023-02-10 19:52:38--  https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/lab03/bola.png
Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...
Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 10267 (10K) [image/png]
Saving to: ‘bola.png’

bola.png            100%[===================>]  10.03K  --.-KB/s    in 0s      

2023-02-10 19:52:38 (80.3 MB/s) - ‘bola.png’ saved [10267/10267]

/content: Scheme missing.
FINISHED --2023-02-10 19:52:38--
Total wall clock time: 0.3s
Downloaded: 1 files, 10K in 0s (80.3 MB/s)
--2023-02-10 19:52:38--  https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/lab03/bolinha.png
Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...
Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 205355 (201K) [image/png]
Saving to: ‘bolinha.png’

bolinha.png         100%[===================>] 200.54K  --.-KB/s    in 0.03s   

2023-02-10 19:52:38 (7.73 MB/s) - ‘bolinha.png’ saved [205355/205355]

/content: Scheme missing.
FINISHED --2023-02-10 19:52:38--
Total wall clock time: 0.2s
Downloaded: 1 files, 201K in 0.03s (7.73 MB/s)
--2023-02-10 19:52:38--  https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/lab03/fuca.png
Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...
Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 46677 (46K) [image/png]
Saving to: ‘fuca.png’

fuca.png            100%[===================>]  45.58K  --.-KB/s    in 0.01s   

2023-02-10 19:52:38 (4.13 MB/s) - ‘fuca.png’ saved [46677/46677]

/content: Scheme missing.
FINISHED --2023-02-10 19:52:38--
Total wall clock time: 0.3s
Downloaded: 1 files, 46K in 0.01s (4.13 MB/s)
--2023-02-10 19:52:39--  https://raw.githubusercontent.com/arnaldojr/cognitivecomputing/master/material/aulas/lab03/bola.png
Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...
Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 10267 (10K) [image/png]
Saving to: ‘bola.png.1’

bola.png.1          100%[===================>]  10.03K  --.-KB/s    in 0s      

2023-02-10 19:52:39 (71.3 MB/s) - ‘bola.png.1’ saved [10267/10267]

/content: Scheme missing.
FINISHED --2023-02-10 19:52:39--
Total wall clock time: 0.05s
Downloaded: 1 files, 10K in 0s (71.3 MB/s)

In [ ]:

img = cv2.imread("fuca.png", cv2.IMREAD_GRAYSCALE)
plt.imshow(img, cmap="Greys_r", vmin=0, vmax=255)

img = cv2.imread("fuca.png", cv2.IMREAD_GRAYSCALE) plt.imshow(img, cmap="Greys_r", vmin=0, vmax=255)

Out[ ]:

<matplotlib.image.AxesImage at 0x7ff0dcc88d90>

In [ ]:

plt.hist(img.ravel(),256,[0,256]); plt.show()

plt.hist(img.ravel(),256,[0,256]); plt.show()

In [ ]:

# normalização de histograma

img_eq = cv2.equalizeHist(img)
plt.imshow(3*img_eq, cmap="Greys_r", vmin=0, vmax=255)

# normalização de histograma img_eq = cv2.equalizeHist(img) plt.imshow(3*img_eq, cmap="Greys_r", vmin=0, vmax=255)

Out[ ]:

<matplotlib.image.AxesImage at 0x252ed7118d0>

In [ ]:

plt.hist(3*img_eq.ravel(),256,[0,256]); plt.show()

plt.hist(3*img_eq.ravel(),256,[0,256]); plt.show()

Podemos fazer o mesmo para uma imgem colorida

In [ ]:

imagem = cv2.imread("bola.png")
image = cv2.cvtColor(imagem, cv2.COLOR_BGR2RGB)
plt.imshow(image, vmin=0, vmax=255); plt.show()
plt.hist(image.ravel(),256,[0,256]); plt.show()

imagem = cv2.imread("bola.png") image = cv2.cvtColor(imagem, cv2.COLOR_BGR2RGB) plt.imshow(image, vmin=0, vmax=255); plt.show() plt.hist(image.ravel(),256,[0,256]); plt.show()

In [ ]:

#histograma Vermelho
plt.imshow(image[:,:,0], cmap="gray", vmin=0, vmax=255); plt.show()
plt.hist(image[:,:,0].ravel(),256,[0,256]); plt.show()

#histograma Vermelho plt.imshow(image[:,:,0], cmap="gray", vmin=0, vmax=255); plt.show() plt.hist(image[:,:,0].ravel(),256,[0,256]); plt.show()

In [ ]:

# Histogrma Verde
plt.imshow(image[:,:,1], cmap="Greys_r", vmin=0, vmax=255); plt.show()
plt.hist(image[:,:,1].ravel(),256,[0,256]); plt.show()

# Histogrma Verde plt.imshow(image[:,:,1], cmap="Greys_r", vmin=0, vmax=255); plt.show() plt.hist(image[:,:,1].ravel(),256,[0,256]); plt.show()

In [ ]:

# Histograma Azul
plt.imshow(image[:,:,2], cmap="Greys_r", vmin=0, vmax=255); plt.show()
plt.hist(image[:,:,2].ravel(),256,[0,256]); plt.show()

# Histograma Azul plt.imshow(image[:,:,2], cmap="Greys_r", vmin=0, vmax=255); plt.show() plt.hist(image[:,:,2].ravel(),256,[0,256]); plt.show()

In [ ]:

image2 = image.copy()
gray_r = image2[:,:,0]
gray_g = image2[:,:,1]
gray_b = image2[:,:,2]

img_bola = image2.copy()

for y in range(0, image2.shape[0]):
    for x in range(0, image2.shape[1]):
        
        if gray_g[y][x] <= 230:
            img_bola[y][x]= 0
        if gray_b[y][x] >= 240:
            img_bola[y][x]= 0    

    

plt.imshow(img_bola, interpolation="none")
plt.show()

image2 = image.copy() gray_r = image2[:,:,0] gray_g = image2[:,:,1] gray_b = image2[:,:,2] img_bola = image2.copy() for y in range(0, image2.shape[0]): for x in range(0, image2.shape[1]): if gray_g[y][x] <= 230: img_bola[y][x]= 0 if gray_b[y][x] >= 240: img_bola[y][x]= 0 plt.imshow(img_bola, interpolation="none") plt.show()

Desafio 1¶

Faça a seguimentação da bolinha de cor laranja. Dica use 2 canais de cores para conseguir seguimentar.

In [ ]:

# Implemente seu código

# Implemente seu código

Desafio 2¶

Faça a seguimentação da bolinha para a imagem "bolinha.png".

In [ ]:

imagem = cv2.imread("bolinha.png")
image = cv2.cvtColor(imagem, cv2.COLOR_BGR2RGB)
plt.imshow(image, vmin=0, vmax=255); plt.show()

imagem = cv2.imread("bolinha.png") image = cv2.cvtColor(imagem, cv2.COLOR_BGR2RGB) plt.imshow(image, vmin=0, vmax=255); plt.show()

In [ ]:

# Implemente seu código

# Implemente seu código

Webcam e opencv¶

Este recurso não vai funcionar no google colab¶

Podemos usar a nossa webcam para registrar imagens e videos, para isso usamos a função cv2.VideoCapture.

In [ ]:

%matplotlib inline
import cv2
from matplotlib import pyplot as plt
import numpy as np
import time as t
print ("OpenCV Version : %s " % cv2.__version__)

%matplotlib inline import cv2 from matplotlib import pyplot as plt import numpy as np import time as t print ("OpenCV Version : %s " % cv2.__version__)

OpenCV Version : 4.5.5 

In [ ]:

webcam = cv2.VideoCapture(0) # pode acontecer do ID da camera não ser 0, ai precisa testar com outros numeros

webcam = cv2.VideoCapture(0) # pode acontecer do ID da camera não ser 0, ai precisa testar com outros numeros

In [ ]:

t.sleep(3) # Espera a webcam ficar pronta

t.sleep(3) # Espera a webcam ficar pronta

In [ ]:

val, image = webcam.read()

val, image = webcam.read()

In [ ]:

val  # Checa se um frame chegou

val # Checa se um frame chegou

Out[ ]:

True

In [ ]:

webcam.release() # fecha a webcam

webcam.release() # fecha a webcam

In [ ]:

plt.imshow(image)

plt.imshow(image)

Out[ ]:

<matplotlib.image.AxesImage at 0x252ed548df0>

Desafio 3¶

Faça uma foto sua.