Skip to content

tutorials
tutorials
Commit 97cffe2f by Usman Shahid
Options

Upload New File

parent 4e1d64d7
Showing with 172 additions and 0 deletions
Querying Azure Custom Vision in Python/Querying an Azure Custom Vision endpoint from Python.ipynb 0 → 100644
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Querying an Azure Custom Vision endpoint from Python\n",
"\n",
"This is the companion notebook to the talk on [Azure Custom Vison for Beginners](https://www.youtube.com/watch?v=OzMRNVolrKE&lc=UgwYIaTgEIkRtacBGI54AaABAg) hosted by [Data Science Dojo](https://datasciencedojo.com/) and delivered by [Usman Shahid](https://www.linkedin.com/in/shahidusmanm/)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import functions as fn\n",
"import nest_asyncio\n",
"nest_asyncio.apply()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#These have already been imported in functions.py. Re-imported them here to show during the webinar.\n",
"\n",
"# Other libraries used\n",
"import os\n",
"\n",
"# for converting the frames into bytes\n",
"import cv2 \n",
"\n",
"# and for processing arrays \n",
"import numpy as np\n",
"\n",
"# for encoding and decoding Custom Vision predictions \n",
"import json\n",
"\n",
"# for converting the Custom Vision predictions to dataframe \n",
"import pandas as pd\n",
"\n",
"# import async packages\n",
"import asyncio\n",
"import aiohttp\n",
"\n",
"# for file name pattern matching \n",
"import fnmatch \n",
"\n",
"# for displaying images from the processes output video \n",
"import matplotlib.pyplot as plt\n",
"\n",
"# importing other required libraries\n",
"import random\n",
"import textwrap\n",
"import datetime \n",
"from PIL import Image\n",
"import time "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#These have already been declared in functions.py. Declared them here to show during the webinar.\n",
"\n",
"# web service end-point for the Custom Vision model \n",
"# we will process video frames (which are images) \n",
"POST_URL = \"Your custom vision endpoint\"\n",
"\n",
"# providing prediction key\n",
"HEADERS = {'Prediction-Key': \"Your custom vision prediction key\", \"Content-Type\":\"application/json\"}\n",
"\n",
"# number of API calls per pool of request \n",
"MAX_CONNECTIONS = 100 \n",
"\n",
"# initializing the height and width for frames in the video \n",
"WIDTH = 0\n",
"HEIGHT = 0"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# creating Output and Stats directories for saving processed videos \n",
"if not os.path.isdir(\"Output\"):\n",
" os.mkdir(\"Output\")\n",
"if not os.path.isdir(\"Stats\"):\n",
" os.mkdir(\"Stats\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# converting raw input video to processed video complete with tags and stats\n",
"threshold=0.3\n",
"fn.ConvertVideo(\"input/input_video.mov\", \"Output/output_video.mp4\", threshold, nframes=70)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# saving processed video as images inside the frames directory\n",
"images = []\n",
"byteImages = []\n",
"vidObj = cv2.VideoCapture(\"Output/output_video.mp4\")\n",
"count = 0\n",
"success = 1\n",
"currentDir = os.getcwd()\n",
"if not os.path.isdir(\"frames\"):\n",
" os.mkdir(\"frames\")\n",
"while success:\n",
" success, image = vidObj.read()\n",
" if success:\n",
" cv2.imwrite(\"frames/frame%d.jpg\" % count, image) \n",
" count += 1\n",
"\n",
"images_num=(len(fnmatch.filter(os.listdir(os.getcwd() + \"/frames\"), '*.jpg')))\n",
"\n",
"# displaying images from frames directory\n",
"i=0\n",
"for i in range(images_num):\n",
" if i%75==0:\n",
" a = plt.imread(\"frames/frame%d.jpg\" % i)\n",
" plt.imshow(a)\n",
" plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.5"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
  • Eric @ericmenk33 commented

    I love how simple ideas can spark the most interesting conversations! It's amazing how a small thought can evolve into something much bigger. 😊 check https://greggsmenusprice.co.uk/

    Edited by Eric
  • Emma @emmabrooks726 commented

    For exploring Cartagena with ease, a hop-on hop-off tour is perfect for discovering the city’s vibrant streets, historic sites, and coastal views at your own pace. cartagena hop on hop off provides travel tips, local insights, and inspiration to make your Cartagena adventure truly memorable. Whether planning your itinerary or catching the latest trends, Zunitive ensures you stay informed and ready to explore.

Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
Explore Help Data Science Dojo