The test output image and the input image are not aligned, and there are crops on the top and left

[kzq666666]

Thanks for your awesome work!
I need to get the edge map corresponding to the original image, but the test output image and the input image are not aligned, and there are crops on the top and left，how to solve it?

[kzq666666]

The test script is as follows

`import numpy as np
import matplotlib.pyplot as plt
import matplotlib.pylab as pylab
import matplotlib.cm as cm
import scipy.misc
from PIL import Image
import scipy.io
import os

import caffe

import sys
caffe_root = '../../' # this file is expected to be in {caffe_root}/examples/hed/

sys.path.insert(0, caffe_root + 'python')

im_lst = []
im = Image.open('../img/0001.png')
im = im.resize((200, 200), Image.ANTIALIAS)
in_ = np.array(im, dtype=np.float32)
in_ = in_[:,:,::-1]
in_ -= np.array((104.00698793,116.66876762,122.67891434))
im_lst.append(in_)

idx = 0
gpu = 3

in_ = im_lst[idx]
in_ = in_.transpose((2,0,1))

caffe.set_mode_gpu()
caffe.set_device(0)

model_root = './'
net = caffe.Net(model_root+'deploy.prototxt', model_root+'hed_pretrained_bsds.caffemodel', caffe.TEST)

net.blobs['data'].reshape(1, 3,200,200)
net.reshape()
net.blobs['data'].data[...] = in_

net.forward()
out1 = net.blobs['sigmoid-dsn1'].data[0][0,:,:]
out2 = net.blobs['sigmoid-dsn2'].data[0][0,:,:]
out3 = net.blobs['sigmoid-dsn3'].data[0][0,:,:]
out4 = net.blobs['sigmoid-dsn4'].data[0][0,:,:]
out5 = net.blobs['sigmoid-dsn5'].data[0][0,:,:]
fuse = net.blobs['sigmoid-fuse'].data[0][0,:,:]

plt.imsave('./test.png', out1)`