Recherche avancée

Sur d’autres sites (5310)

• Why JPEG compressing an uncompressed image differs its original (FFmpeg, NvJPEG, ...)

  22 juin 2021, par Fruchtzwerg

  I am currently struggling to understand why recompressing an uncompressed JPEG image differs its original.

  


  It's clear, that JPEG is a lossy compression, but what if the image to compress is already uncompressed, which means all sampling losses are already included ? In other words : Downsampling and DCT should be inversable at this point without loosing data.

  


  

  


  To make sure losses are not effected by the color space conversion, this step is skipped and YUV images are used.

  


  

  1. Compress YUV image to JPEG (image.yuv —> image.yuv.jpg)

  


  2. Uncompress JPEG image to YUV image (image.yuv.jpg —> image.yuv.jpg.yuv)

  


  3. Compress YUV image to JPEG (image.yuv.jpg.yuv —> image.yuv.jpg.yuv.jpg)

  


  4. Uncompress JPEG image to YUV image (image.yuv.jpg.yuv.jpg —> image.yuv.jpg.yuv.jpg.yuv)

  


  


  Step 1 includes a lossy compression, so we will not deal with this step anymore. For me, intresting is what happens afterwards :

  


  Uncompressing the JPEG image back to YUV (step 2) leads to an image which perfectly fits all sampling steps if compressed again (step 3). So the JPEG image after step 3 should (from my understanding) be exactly the same as after step 1. Also the YUV images after step 4 and step 2 should equal each other.

  


  Looking at the steps for one 8x8 block the following simplified sequence should illustrate what I am trying to descibe. Lets start with the original YUV image, which can only be decompressed loosing all decimal places :

  


  [ 1.123, 2.345, 3.456, ... ]    (YUV)
    DTC + Quantization
[ -26, -3, -6, ... ]            (Quantized frequency space)
    Inverse DTC + Quantization
[ 1, 2, 3, ... ]                (YUV)


  


  Doing this with input, which already matches all steps, which may lead to loss of data afterwards (using round numbers in my example), the decompressed image should match its original :

  


  [ 1, 2, 3, ... ]                (YUV)
    DTC + Quantization
[ -26, -3, -6, ... ]            (Quantized frequency space)
    Inverse DTC + Quantization
[ 1, 2, 3, ... ]                (YUV)


  


  There are also some sources and discussions, which are confirming my idea :

  


  

  • need help creating Jpeg Generational Degradation code

  


  • What factors cause or prevent “generational loss” when JPEGs are recompressed multiple times ?

  


  • Lossless Chroma Subampling

  


  


  So much for theory. In praxis, I've runned these steps using ffmpeg and Nvidias jpeg samples (using NvJPEGEncoder).

  


  ffmpeg :

  


  #Create YUV image
ffmpeg -y -i image.jpg -s 1920x1080 -pix_fmt yuv420p image.yuv
#YUV to JPEG
ffmpeg -y -s 1920x1080 -pix_fmt yuv420p -i image.yuv image.yuv.jpg
#JPEG TO YUV
ffmpeg -y -i image.yuv.jpg -s 1920x1080 -pix_fmt yuv420p image.yuv.jpg.yuv
#YUV to JPEG
ffmpeg -y -s 1920x1080 -pix_fmt yuv420p -i image.yuv.jpg.yuv image.yuv.jpg.yuv.jpg
#JPEG TO YUV
ffmpeg -y -i image.yuv.jpg.yuv.jpg -s 1920x1080 -pix_fmt yuv420p image.yuv.jpg.yuv.jpg.yuv
#YUV to JPEG
ffmpeg -y -s 1920x1080 -pix_fmt yuv420p -i image.yuv.jpg.yuv.jpg.yuv image.yuv.jpg.yuv.jpg.yuv.jpg


  


  Nvidia :

  


  #Create YUV image
./jpeg_decode num_files 1 image.jpg image.yuv
#YUV to JPEG
./jpeg_encode image.yuv 1920 1080 image.yuv.jpg
#JPEG TO YUV
./jpeg_decode num_files 1 image.yuv.jpg image.yuv.jpg.yuv
#YUV to JPEG
./jpeg_encode image.yuv.jpg.yuv 1920 1080 image.yuv.jpg.yuv.jpg
#JPEG TO YUV
./jpeg_decode num_files 1 image.yuv.jpg.yuv.jpg image.yuv.jpg.yuv.jpg.yuv
#YUV to JPEG
./jpeg_encode image.yuv.jpg.yuv.jpg.yuv 1920 1080 image.yuv.jpg.yuv.jpg.yuv.jpg


  


  But a comparison of the images

  


  

  • image.yuv.jpg.yuv and image.yuv.jpg.yuv.jpg.yuv

  


  • image.yuv.jpg.yuv.jpg and image.yuv.jpg.yuv.jpg.yuv.jpg

  


  


  showing differences in the files. That brings me to my question why and where the difference gets happen, since from my understanding the files should be equal.

  


• Multiple files concatenation using ffmpeg by the use of GPU Hardware acceleration

  30 juin 2021, par bubus

  I need to concatenate multiple mp4, h264 encoded files into single one together with speed up filter, using GPU HW acceleration. I am using Debian 10 Buster 64bit, and the card I am using is Nvidia Gainward GTX960.

  


  I have installed CUDA, together with Nvidia driver and configured ffmpeg with the following parameters :

  


  ./configure --enable-nonfree -–enable-cuda-sdk –enable-libnpp --extra-cflags=-I/usr/local/cuda/include --extra-ldflags=-L/usr/local/cuda/lib64`


  


  The problem is, I believe, that GPU is not working at 100%, so the concatenation takes quite a long time.

  


  The command I am using to concatenate files and speeding them up :

  


  ./ffmpeg -c:v h264_cuvid -f concat -i mylist.txt -c:a copy -c:v h264_nvenc -y -filter:v 'setpts='0.0625'*PTS' -an merged.mp4


  


  Output of nvidia-smi while the above command is being executed :

  


  

  


  ffmpeg version: ffmpeg version N-102801-gb74beba9a9
CUDA version: 11.2
Nvidia driver version: 460.84


  


  I have no idea what else can I do to speed up the concatenation.

  


• How to save frames to memory from video with ffmpeg gpu encoding ?

  6 juillet 2021, par Spetra

  I'm trying to extract frames from video and save them to memory(ram).
With CPU encoding, I don't have any problems :

  


  ffmpeg -i input -s 224x224 -pix_fmt bgr24 -vcodec rawvideo -an -sn -f image2pipe -


  


  But when I'm trying to use some NVIDIA GPU encoding, I'm always getting noisy images.
I tried to use different commands, but the result was always the same, on Windows and Ubuntu.

  


  ffmpeg -hwaccel cuda -i 12.mp4 -s 224x224 -f image2pipe - -vcodec rawvideo


  


  With saving JPG's on disk, I don't have any problems.

  


  ffmpeg -hwaccel cuvid -c:v h264_cuvid -resize 224x224 -i {input_video} \
     -vf thumbnail_cuda=2,hwdownload,format=nv12 {output_dir}/%d.jpg


  


  There was my python code for testing these commands :

  


  import cv2
import subprocess as sp
import numpy

IMG_W = 224
IMG_H = 224
input = '12.mp4'

ffmpeg_cmd = [ 'ffmpeg','-i', input,'-s', '224x224','-pix_fmt', 'bgr24',  '-vcodec', 'rawvideo', '-an','-sn', '-f', 'image2pipe', '-']


#ffmpeg_cmd = ['ffmpeg','-hwaccel' ,'cuda' ,'-i' ,'12.mp4','-s', '224x224','-f' , 'image2pipe' ,'-' , '-vcodec' ,'rawvideo']

pipe = sp.Popen(ffmpeg_cmd, stdout = sp.PIPE, bufsize=10)
images = []
encode_param = [int(cv2.IMWRITE_JPEG_QUALITY), 95]
cnt = 0
while True:
    cnt += 1
    raw_image = pipe.stdout.read(IMG_W*IMG_H*3)
    image =  numpy.fromstring(raw_image, dtype='uint8')     # convert read bytes to np
    if image.shape[0] == 0:
        del images
        break   
    else:
        image = image.reshape((IMG_H,IMG_W,3))
        

    cv2.imshow('test',image)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

pipe.stdout.flush()
cv2.destroyAllWindows()