pixelconversion.py

# -*- coding: utf-8 -*-
"""pixelconversion.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1hfdgkYOw8w6DdJqsZx8txmw8INUGxesl
"""


# pip install pymupdf -q

"""### Imports"""

import fitz
from PIL import Image
import numpy as np
import cv2
import db
import tsadropboxretrieval
import fitz
from io import BytesIO

################################################
### Check if page is visually rotated or not 

def is_content_rotated(mask):
      contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
      # Check the orientation of the purple shape 
      for contour in contours:
          rect = cv2.minAreaRect(contour)
          angle = rect[-1]
          print(angle)
          if angle != 0.0 and angle != 90.0 and angle != 180.0:
              print("Page content appears visually rotated.")
              return True
      print("Page content does not appear visually rotated.")
      return False

def drawisrotated(data):
  docPath = fitz.open("pdf",data) #dropbox path
  pageDocPath=docPath[0]
  if pageDocPath.rotation !=0:
   pageDocPath.draw_rect([250,250,pageDocPath.mediabox.width-500,pageDocPath.mediabox.height-500],  color = (75/255,0,130/255), width = 1,fill=(75/255,0,130/255),fill_opacity=0.9)
  else:
   pageDocPath.draw_rect([0+250,0+250,pageDocPath.mediabox.width-500,pageDocPath.mediabox.height-500],  color =(75/255,0,130/255), width = 1,fill=(75/255,0,130/255),fill_opacity=0.9)
  
  pix = pageDocPath.get_pixmap()
  image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
  img = cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)
  mask=DetectColor(img,color=(75,0,130))
  bool_T_F=is_content_rotated(mask)
  print(bool_T_F)
  if bool_T_F==True:
    if pageDocPath.mediabox.height >pageDocPath.mediabox.width:
      rectText=fitz.Rect(300, 200, (pageDocPath.mediabox.width-80),( pageDocPath.mediabox.width+80) )
    else:
      rectText=fitz.Rect(500, 200, (pageDocPath.mediabox.height-80),( pageDocPath.mediabox.height+80) )

    text = """Scale Document"""
    annot1=pageDocPath.add_freetext_annot(rectText, text, fontsize=45, fontname='helv', border_color=(1,1,1), text_color=(1,1,1), rotate= pageDocPath.rotation, align=1)
    annot1.update()
    return docPath
  else:
    doc=openDrawPDF(data)
    return doc


"""### Open PDF and draw a rectangle on it using Fitz"""
def openDrawPDF(data):
  
  doc=fitz.open("pdf", data)
  page = doc[0]
  
  if page.rect.height > page.rect.width:
    rectText=fitz.Rect(300, 200, (page.rect.width-80),( page.rect.width+80) )
  else:
    rectText=fitz.Rect(500, 200, (page.rect.height-80),( page.rect.height+80) )
  if page.rotation !=0:
    page.draw_rect([0+10,0+10,page.rect.height-10,page.rect.width-10],  color = (75/255,0,130/255), width = 1,fill=(75/255,0,130/255),fill_opacity=0.9 )
  else:
    page.draw_rect([0+10,0+10,page.rect.width-10, page.rect.height-10],  color = (75/255,0,130/255), width = 1,fill=(75/255,0,130/255),fill_opacity=0.9 )
 
  text = """Scale Document"""
  annot1=page.add_freetext_annot(rectText, text, fontsize=45, fontname='helv', border_color=(1,1,1), text_color=(1,1,1), rotate= page.rotation, align=1)
  annot1.update()

  return doc#,encrypt,perm
"""### Extract color"""

def DetectColor(img,color=0):

  imgCopy=img.copy()
  imgCopy=cv2.cvtColor(imgCopy,cv2.COLOR_BGR2HSV)
  tol=5 #tolerance
  # color=hexRGB(color)
  h,s,v = cv2.cvtColor(np.uint8([[[color[2],color[1],color[0]]]]),cv2.COLOR_BGR2HSV)[0][0]

  lower =np.array( [h- tol, 100, 100 ], dtype='uint8')
  upper = np.array( [h + tol, 255, 255],dtype='uint8')

  mask = cv2.inRange(imgCopy, lower , upper)

  detectedColors = cv2.bitwise_and(imgCopy,imgCopy, mask= mask)   # Bitwise-AND mask and original image

  kernel=np.ones((3,3),np.uint8)
  mask=cv2.dilate(mask,kernel, iterations=5)
  mask=cv2.erode(mask,kernel, iterations=4)

  detectedColors=cv2.dilate(detectedColors,kernel, iterations=5)
  detectedColors=cv2.erode(detectedColors,kernel, iterations=4)

  detectedColors=cv2.cvtColor(detectedColors,cv2.COLOR_HSV2BGR)
  detectedColors=cv2.medianBlur(detectedColors,7)
  # cv2_imshow(detectedColors)

  return mask

"""### For backend - calc area and perim"""

def getAreaPerimeter(path,name):
  dbxTeam=tsadropboxretrieval.ADR_Access_DropboxTeam('user')
  md, res =dbxTeam.files_download(path= path+name)
  data = res.content
  doc=fitz.open("pdf", data)
  area=0
  perimeter=0
  for page in doc:
      pix = page.get_pixmap(dpi=300)  # render page to an image
      pl=Image.frombytes('RGB', [pix.width,pix.height],pix.samples)
      img=np.array(pl)
      print(img.shape)
      img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
  mask=DetectColor(img,color=(73,0,130)) #detect colored rect drawn on the pdf
  # cv2.imwrite('maskk.png',mask)
  contours, hierarchy = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
  for contour in contours:
      area = cv2.contourArea(contour)
      perimeter = cv2.arcLength(contour, True)
  return area,perimeter