app.py

import streamlit as st  
import base64
import os.path  
import os
import cv2  
import random
   
os.system("ls")
col1, col2 = st.columns(2)

title = "Melhoria de imagens"
st.title(title)
os.system("ls")
description = "Sistema para automação。"
st.header(description)
article = "<p style='text-align: center'><a href='https://huggingface.co/spaces/akhaliq/GFPGAN/' target='_blank'>clone from akhaliq@huggingface with little change</a> | <a href='https://github.com/TencentARC/GFPGAN' target='_blank'>GFPGAN Github Repo</a></p><center><img src='https://visitor-badge.glitch.me/badge?page_id=akhaliq_GFPGAN' alt='visitor badge'></center>"


uploaded_file = st.file_uploader("Choose a file")
if uploaded_file is not None:
    # To read file as bytes:
    bytes_data = uploaded_file.getvalue()
    st.video(bytes_data)
    os.system("rm -R /tmp/*")
    os.system("mkdir /tmp/_input")
    os.system("mkdir /tmp/_output")
    os.system("mkdir /tmp/_outputf")
    os.system("mkdir /tmp/videoSaida")
    with open("/tmp/inputvideo", "wb") as binary_file:
       binary_file.write(bytes_data)
    vcap = cv2.VideoCapture('/tmp/inputvideo') # 0=camera
    width=0
    height=0
    if vcap.isOpened(): 
              # get vcap property 
              width  = vcap.get(3)   # float `width`
              height = vcap.get(4)  # float `height`
              
              st.write(str(width)+"x"+str(height))
              fps = vcap.get(5)
              st.write(str(fps)+" fps")
@st.experimental_memo(suppress_st_warning=True)
def chamada():
          

          percent_complete=0
          my_bar = st.progress(0)   
          percent_complete= percent_complete+ 10
          with col1:
           my_bar.progress(percent_complete )    
       #exec=True 
#           st.write("ffmpeg separando imagens")
       #if not os.path.isfile("./_input/imagem-0001.png"):
          vcap = cv2.VideoCapture('/tmp/inputvideo') # 0=camera
          width=0
          height=0
          if vcap.isOpened(): 
              # get vcap property 
              width  = vcap.get(3)   # float `width`
              height = vcap.get(4)  # float `height`
              fps = vcap.get(5)

        #  os.system("ffmpeg -i /tmp/inputvideo -vn -acodec copy /tmp/inputvideo.acc")
          os.system("ffmpeg -i /tmp/inputvideo -compression_level 10 -pred mixed -pix_fmt rgb24 -sws_flags +accurate_rnd+full_chroma_int -s "+str(int(width/2))+"x"+str(int(height/2))+" -r   29.97 /tmp/_input/imagem-%4d.png")
          percent_complete= percent_complete+ 30
          with col1:
           my_bar.progress(percent_complete )   
 #          st.write("testando imagem")
          input_img = cv2.imread("/tmp/_input/imagem-0002.png" , cv2.IMREAD_COLOR)
          input_img= cv2.cvtColor(input_img,cv2.COLOR_BGR2RGB) 
          with col1:
           st.image(input_img)
          
          os.system("ls /tmp/_input") 
          if 'myVar' not in globals():
              myVar=""
  #        st.write("melhorando faces")
          with col2:
           with st.spinner('Wait for it...'):
             # os.system("pip install git+https://github.com/TencentARC/GFPGAN.git")
                    os.system("python3 inference_gfpgan.py -i /tmp/_input -o /tmp/_output -v 1.4 -s 2")
          with col1:
           percent_complete= percent_complete+ 30
           my_bar.progress(percent_complete )   
         
          os.system("ls /tmp/_output")
          os.system("echo ----")
          os.system("ls /tmp/_output/cmp")
          os.system("echo ----")
          os.system("ls /tmp/_output/restored_imgs")
          os.system("echo ----")
#          s 1480x2560
       #   with col1:
   #        st.write("recompilando video")
          #ffmpeg -r 60 -f image2 -s 1920x1080 -i _output/restored_imgs/imagem-%4d.png   -pix_fmt yuv420p ./videoSaida/output.mp4
          os.system("ffmpeg -y -r 29.97 -f image2  -i /tmp/_output/restored_imgs/imagem-%4d.png   -pix_fmt yuv420p /tmp/videoSaida/output.mp4")
         # os.system("ffmpeg -i /tmp/videoSaida/outputp1.mp4 -i /tmp/inputvideo.acc -c:v copy -c:a aac /tmp/output.mp4")
          os.system("ls /tmp/videoSaida")
          #st.video("./videoSaida/output.mp4" )
          with col1:
        #   st.write("preparando para download do video")
           percent_complete= percent_complete+ 30
           my_bar.progress(percent_complete ) 
          with col2:
           with open("/tmp/videoSaida/output.mp4", "rb") as file:
              st.video(file)

              btn = st.download_button(
          
                      label="Download video",
          
                      data=file,
          
                      file_name="output.mp4",
          
                      mime="video/mp4"
          
                    )
          #st.download_button("download video", data, file_name='output.mp4', mime='video/mp4',)
          #stremio
          #input_img = cv2.imread("./_output/cmp/imagem-0001_0000.png" , cv2.IMREAD_COLOR)
          #input_img = cv2.imread("./_output/cmp/imagem-0001_0000.png" , cv2.IMREAD_COLOR)
          with col2:
         #  st.write("demonstrando imagem restaurada")
           input_img = cv2.imread("/tmp/_output/restored_imgs/imagem-0002.png" , cv2.IMREAD_COLOR)
           input_img= cv2.cvtColor(input_img,cv2.COLOR_BGR2RGB) 
          
           st.image(input_img)
          exec=False
       #return Image.fromarray(restored_faces[0][:,:,::-1])
load = st.checkbox("load")

if load:
       chamada()
   
#st.button('Imagem',on_click=inference)
exec=True