CODEWITHSUNDEEP
pythonkerasdeep-learning
codcod4
codcod4

Reputation: 61

How can I use Keras with webcam?

I have an pretrained model. Model trained with 20000 "gray" sample. İt is working with "gray" test samples. But I want to test this model with webcam. Here my code:

#Load the saved model
model = keras.models.load_model('C:\keras\handrecognition_model.h5')
video = cv2.VideoCapture(0)

while True:
    _, frame = video.read()
    im = Image.fromarray(frame, 'RGB')
    im = im.resize((128, 128))
    img_array = np.array(im)

    img_array = np.expand_dims(img_array, axis=0)

    prediction = int(model.predict(img_array)[0][0])

    # if prediction is 0, which means I am missing on the image, then show the frame in gray color.
    if prediction == 0:
            frame = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

    cv2.imshow("Capturing", frame)
    key = cv2.waitKey(1)
    if key == ord('q'):
            break

video.release()
cv2.destroyAllWindows()

There is an error : ValueError: Error when checking input: expected conv2d_1_input to have shape (120, 320, 1) but got array with shape (128, 128, 3).

Here output of gray test images: enter image description here

Model trained:

# Construction of model
model = Sequential()
model.add(Conv2D(32, (5, 5), activation='relu', input_shape=(120, 320, 1))) 
model.add(MaxPooling2D((2, 2)))
model.add(Conv2D(64, (3, 3), activation='relu')) 
model.add(MaxPooling2D((2, 2)))
model.add(Conv2D(64, (3, 3), activation='relu'))
model.add(MaxPooling2D((2, 2)))
model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dense(10, activation='softmax'))

Edit: I update code like this:

_, frame = video.read()
frame = cv2.resize(frame, (120, 360))
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
img_array = np.array(gray)

ValueError: Error when checking input: expected conv2d_1_input to have 4 dimensions, but got array with shape (1, 360, 120)

Edit 2: Here train article: https://towardsdatascience.com/tutorial-using-deep-learning-and-cnns-to-make-a-hand-gesture-recognition-model-371770b63a51

Edit 3: I guess, it is working. Now I will send frame to predict and I will found hand gest. I will share if I can do. Thank you.

    _, frame = video.read()
    frameCopy=frame.copy()
    frameCopy = cv2.resize(frameCopy, (120, 320))
    gray = cv2.cvtColor(frameCopy, cv2.COLOR_BGR2GRAY)
    img_array = np.array(gray)
    img_array = img_array.reshape(120, 320, 1)
    img_array = np.expand_dims(img_array, axis=0)

Upvotes: 3

Views: 1195

Answers (2)

Theodor Peifer
Theodor Peifer

Reputation: 3496

Answer for you question after your edit: You need 4 dimensions not three: (batch-size, channels, width, height). So try the following:

img_array = np.array(gray)
img_array = img_array.reshape(1, 1, 360, 120)

Upvotes: 1

theastronomist
theastronomist

Reputation: 1056

    im = Image.fromarray(frame, 'RGB')
    im = im.resize((128, 128))
    im = im.convert('RGB')

This will convert your grayscale image to RGB. It will just copy the values from the one dimension you currently have to 3 dimensions. If it throws an error, try this instead

    im = im.convert('L')
    im = im.convert('RGB')

'L' converts it to black and white first in case your input is sometimes not just 1D

Upvotes: 0

Related Questions