首页 > 解决方案 > 为什么我在 np.array 时出现重塑错误(图片)

问题描述

所以我正在做一个手写识别项目,它使用 IamDB 可以正常工作。但是,当我尝试自己拍摄手写照片时,出现此错误,知道如何解决吗?我尝试将图片更改为灰度,但它不起作用

images = []
PTH = Path_to_test 
#filename  = os.path.basename(PTH[0])
filename = PTH[0]
#******************************************************
print (filename)
im = Image.open(filename)
#print ("cu")
cur_width = im.size[0]
cur_height = im.size[1]
    # print(cur_width, cur_height)
​
height_fac = 113 / cur_height
​
new_width = int(cur_width * height_fac)
size = new_width, 113
​
imresize = im.resize((size), Image.ANTIALIAS)  # Resize so height = 113 while keeping aspect ratio
now_width = imresize.size[0]
now_height = imresize.size[1]
    # Generate crops of size 113x113 from this resized image and keep random 10% of crops
​
avail_x_points = list(range(0, now_width - 113 ))# total x start points are from 0 to width -113
​
    # Pick random x%
factor = 0.1 
pick_num = int(len(avail_x_points)*factor)

random_startx = sample(avail_x_points,  pick_num)
​
for start in random_startx:
  imcrop = imresize.crop((start, 0, start+113, 113))
  images.append(np.asarray(imcrop))
T_test = np.array(images)
print (T_test.shape)
T_test = T_test.reshape(T_test.shape[0], 113, 113, 1)
    #convert to float and normalize
T_test = T_test.astype('float32')
T_test /= 255
shuffle(T_test)  
print (T_test.shape)

T_test shape 为 (3, 113, 113, 3),是一张黑白的手写图片。这是错误:

     34 T_test = np.array(images)
     35 print (T_test.shape)
---> 36 T_test = T_test.reshape(T_test.shape[0], 113, 113, 1)
     37     #convert to float and normalize
     38 T_test = T_test.astype('float32')

ValueError:无法将大小为 114921 的数组重塑为形状 (3,113,113,1) 我也使用

predictions = model.predict(T_test, verbose =1)

所以我不能把它改成 (T_test.shape[0], 113, 113, 3)

标签: pythonmachine-learningpython-imaging-library

解决方案


图像有 3 个通道,因此您需要使用 3 而不是 1:

T_test = T_test.reshape(T_test.shape[0], 113, 113, 3)

推荐阅读