Question

我正在研究猫与狗的分类模型，以加载以下数据。我要做的是使用OpenCV读取图像，然后使用边缘检测获取边缘并将图像的大小调整为200,200，因此最后我得到了黑白图像，白色是检测到的轮廓。

import os
import sys
import cv2

import random

from tqdm import tqdm

import matplotlib.pyplot as plt

train_images = os.listdir('data/train')
test_images = os.listdir('data/test')

test_images_data = []

for image in tqdm(test_images):
    image_data = cv2.imread('data/test/' + image)

    #Convert to GrayScale
    #gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

    #convert color from BGR to RGB
    image_data = cv2.cvtColor(image_data, cv2.COLOR_BGR2RGB)
    image_data = cv2.resize(image_data, (200, 200))

    #turn to only borders
    edges = cv2.Canny(image_data, 150, 150)

    test_images_data.append(edges)



train_images_data = []
train_images_labels = []

random.shuffle(train_images)

for image in tqdm(train_images):
    image_data = cv2.imread('data/train/' + image)

    #Convert to GrayScale
    #gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

    #convert color from BGR to RGB
    image_data = cv2.cvtColor(image_data, cv2.COLOR_BGR2RGB)
    image_data = cv2.resize(image_data, (200, 200))

    #turn to only borders
    edges = cv2.Canny(image_data, 150, 150)

    train_images_data.append(edges)

    if image.startswith('cat'):
        train_images_labels.append(0)
    else:
        train_images_labels.append(1)



from keras.models import Sequential 
from keras.layers import Conv2D, MaxPooling2D 
from keras.layers import Activation, Dropout, Flatten, Dense


model = Sequential() 
model.add(Conv2D(32, (2, 2), input_shape=train_images_data[0].shape)) 
model.add(Activation('relu')) 
model.add(MaxPooling2D(pool_size=(2, 2))) 

model.add(Conv2D(32, (2, 2))) 
model.add(Activation('relu')) 
model.add(MaxPooling2D(pool_size=(2, 2))) 

model.add(Conv2D(64, (2, 2))) 
model.add(Activation('relu')) 
model.add(MaxPooling2D(pool_size=(2, 2))) 

model.add(Flatten()) 
model.add(Dense(64)) 
model.add(Activation('relu')) 
model.add(Dropout(0.5)) 
model.add(Dense(1)) 
model.add(Activation('sigmoid')) 

model.compile(loss='binary_crossentropy', 
            optimizer='rmsprop', 
            metrics=['accuracy']) 

model.fit(train_images_data, train_images_labels, epochs=150, validation_split=0.2, batch_size=10)

问题是当我像这样运行它时出现此错误

ValueError: Input 0 is incompatible with layer conv2d_16: expected ndim=4, found ndim=3

我尝试了其他input_shape值，例如

(200, 200, 1)
(200, 200, -1)
(1, 200, 200)
(-1, 200, 200)

没有工作。

Answer 1

输入形状：（200，200，1，），以使批次大小为最后一个尺寸，因此预期为4dim。

Answer 2

问题有两个方面。首先，我需要将数据从列表转换为numpy数组，然后对其进行整形。

train_images_data = np.array(train_images_data)
train_images_data = train_images_data.reshape([-1, 200, 200,1])

现在这可以了

model.add(Conv2D(32, (3, 3), input_shape=(200, 200, 1)))

Keras输入一维图像数组，指定正确的输入形状

2 个答案: