TersorFlow如何预期展平图像？

Question

当使用conv_2d层TersorFlow执行二维卷积，它期望的像素上排列为

[ 
    [img[i].red, img[i].green, img[i].blue], 
    [img[i+1].red, etc.], 
] 

或者

[ 
    [mg[i].red, img[i+1].red, etc.], 
    [mg[i].green, img[i+1].green, etc.], 
] 

或一些其他的方式？

Answer 1

2D卷积期望的4张量d为具有以下形状输入：

[batch_size时，IMAGE_HEIGHT，IMAGE_WIDTH，channel_size]

在RGB图像的情况下，信道是三种颜色。因此，像素应当排列为：

[ 
    [ 
    [img[i,j].red, img[i,j].green, img[i,j].blue], 
    [img[i, j+1].red, img[i, j+1].green, img[i, j+1].blue], 
    etc 
    ], 
    [ 
    [img[i+1,j].red, img[i+1,j].green, img[i+1,j].blue], 
    [img[i+1, j+1].red, img[i+1, j+1].green, img[i+1, j+1].blue], 
    etc 
    ], 
    etc 
] 

（与img[y_coordinate, x_coordinate]和img[i,j] = img[i*image_width + j]）