Skip to content

finish the script to transform matrix to png image for debugging #1575

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
danpovey merged 1 commit into from
Apr 25, 2017
Merged

finish the script to transform matrix to png image for debugging #1575

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
63 changes: 53 additions & 10 deletions egs/cifar/v1/image/matrix_to_image.py
100644 → 100755
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,36 +5,79 @@
# Apache 2.0


""" This script converts a Kaldi-format text matrix into a jpeg image.
It reads the matrix from its stdin and writes the jpeg image to its
""" This script converts a Kaldi-format text matrix into a png image.
It reads the matrix from its stdin and writes the png image to its
stdout.
For instance:
cat <<EOF | image/matrix_to_image.py --color=true > foo.jpeg
cat <<EOF | image/matrix_to_image.py --color 3 > foo.png
[ 0.0 0.5 1.0
0.0 0.0 0.0 ]
EOF
The image format is that the number of rows equals the width of the image, and the
number of columns equals the height of the image times the number of channels
(1 for black and white, 3 for color (RGB)), with the channel varying the
fastest. The above example would produce a color image with width 2 and
height 1.

height 1. The first row corresponds to the left side of the image, and the
first column corresponds to the top of the image.
"""

import argparse
import os
import sys
import numpy as ny
from PIL import Image


parser = argparse.ArgumentParser(description="""Converts Kaldi-format text matrix
representing an image on stdin into jpeg image on stdout. See
representing an image on stdin into png image on stdout. See
comments at top of script for more details.""")

parser.add_argument('--color', type=bool, default=True,
help='True if the image is in color ')
parser.add_argument('--color', type=int, default=3,
help='3 if the image is in RGB, 1 if the image is in grayscale ')


args = parser.parse_args()

matrix = []
num_rows = 0
num_cols = 0
while True:
tmp = sys.stdin.readline().strip('\n').split()
if tmp == []:
break
if tmp[0] == '[': # drop the "[" in the first row
tmp = tmp[1:]
if tmp[-1] == ']': # drop the "]" in the last row
tmp = tmp[:-1]
if num_rows == 0:
num_cols = len(tmp) # initialize
if len(tmp) != num_cols:
raise Exception("All rows should be of same length")
tmp = map(float, tmp) # string to float
if max(tmp) > 1:
raise Excetion("Elmement vaule in the matrix should be normalized and no larger than 1")
tmp = [int(x * 255) for x in tmp] # float to integer ranging from 0 to 255
matrix.append(tmp)
num_rows+=1

if args.color == 3:
if num_cols%3!=0:
raise Exception("Number of columns should be 3*n in the colorful mode")
width = num_rows
height = num_cols/3

image_array = ny.zeros((height, width, chan), dtype=ny.uint8)
for i in range(height):
for j in range(width):
image_array[i,j] = [matrix[j][3*i], matrix[j][3*i+1], matrix[j][3*i+2]]
im = Image.fromarray(image_array)
im.save(sys.stdout,'png')
else:
width = num_rows
height = num_cols
image_array = ny.zeros((height,width),dtype=ny.uint8)
for i in range(height):
for j in range(width):
image_array[i,j] = matrix[j][i]
im = Image.fromarray(image_array)
im.save(sys.stdout,'png')

# TODO.