在 Python 上读取并转换 Heightmap.RAW 文件（灰度 16 位）

Question

I have a grayscale 16bit heightmap from a game that i want to open and convert to PNG/JPEG using python, i've tried a lot of ways, found some tools, but still weren't able to do that i want...

And i just want to say, that i want to do this, because each map of that game has around 30+ heightmaps files that are later align in a grid, and basically i want to do the same the game does, put every file in a grid and then export it as a single file (PNG or JPEG), i was able to do that with the minimaps, which are.DDS files, so converting was easy, and now i need to do the same for the heightmaps.

Now, what i've tried so far...

1: (from this how to convert raw images to png in python? )

rawData = open("foo.raw" 'rb').read() imgSize = (x,y)
# Use the PIL raw decoder to read the data.
# the 'F;16' informs the raw decoder that we are reading 
# a little endian, unsigned integer 16 bit data. 
img = Image.fromstring('L', imgSize, rawData, 'raw', 'F;16') img.save("foo.png") 

On this the Image.fromstring() was giving an error saying that fromstring doesn't exist, and i couldn't find a way to replace it

2: Rawpy library examples

import rawpy
import imageio

path = 'image.raw'
with rawpy.imread(path) as raw:
    rgb = raw.postprocess()
imageio.imsave('default.tiff', rgb)

On this, rawpy just wasn't able to read the file, saying it was a raw file

3: OpenCV library

import cv2
import numpy as np
from PIL import Image
import torchvision

with open('height.raw', 'rb') as infile:
     buf = infile.read()

x = np.fromstring(buf, dtype='uint8')

img = cv2.imdecode(x, cv2.IMREAD_UNCHANGED)

image = torchvision.transforms.ToPILImage(img)
image = Image.open(image)
image.show()

Can't make it show/export the image, even if opencv was able to open it correctly, i still couldn't figure out a way.

And more and more stuff...

But with all this search i found out some things that helped but still wasn't enough

First, i found out that i was able to insert the heightmap on photoshop and make it look as i liked by importing it with 131x131 resolution, 16 bit, PC IBM, but again, this needs to be automated since they are over 30+ files per map, and over 20 maps... Then, i found out a website called rawpixels.net that i was also able to load the image, properly by changing some settings:

width: 131 height: 131 offset: 1 Predefined Format: Grayscale 8bit bpp1: 16

But again, automating it, the website uses javascript for reading the image and changing the settings, so converting this to python isn't that easy, what i tried after founding out about this website was, using js2py library run the functions on the page (no luck), trying to convert the javascript code to a python code (also no luck), and this was all yesterday, today on the work, i was thinking, maybe i can save the page as a single.html file, editing the.html file to have the default settings as i want, and then through python trying to load the.html file and make it load the map file, and then using js2py searching for the result div image, and getting it from there, didn't try this tho, but seems to be a bit complex, at least for me. ..

Sorry for having such a big post... Anyways, thks in advance!

Answer 1

UPDATE:

import cv2
import numpy as np

width = 131
height = 131

with open("height.Raw", "rb") as rawimg:
   img = np.fromfile(rawimg, np.dtype('u2'), width * height).reshape(height, width)
   colimg = cv2.cvtColor(img, cv2.IMREAD_GRAYSCALE)
   cv2.imwrite("test.png", colimg)

Got it to work!!! Thks to who made this https://www.pythonfixing.com/2022/01/fixed-reading-and-saving-12bit-raw.html