Jeff Hobbs jeffehobbs

Using Python 3 + Google Cloud Vision API's OCR to extract text from photos and scanned documents

Just a quickie test in Python 3 (using Requests) to see if Google Cloud Vision can be used to effectively OCR a scanned data table and preserve its structure, in the way that products such as ABBYY FineReader can OCR an image and provide Excel-ready output.

The short answer: No. While Cloud Vision provides bounding polygon coordinates in its output, it doesn't provide it at the word or region level, which would be needed to then calculate the data delimiters.

On the other hand, the OCR quality is pretty good, if you just need to identify text anywhere in an image, without regards to its physical coordinates. I've included two examples:

####### 1. A low-resolution photo of road signs

Q: How do I combine individual video clips into one movie file for upload to YouTube using ffmpeg?

The following assumes you have ffmpeg installed on your Mac. If you need to install it, please use Homebrew.

The settings for encoding for YouTube with ffmpeg can be found here and here.

There is no error checking here. It assumes that there are videos in the folder, and that they have mp4 extensions.

	var GetIpInfo = function(ipAddr) {
	var info = null;
	var infoUrl = "http://ipinfo.io/" + ipAddr;
	$.ajax({
	url: infoUrl,
	type: 'GET',
	dataType: 'json',
	async: false,
	success: function(data) {
	info = data;

	#!/usr/bin/env python
	import argparse
	import logging
	try:
	from urllib import splittype
	except ImportError:
	from urllib.parse import splittype

	import boto3

	from selenium import webdriver
	from selenium.webdriver import DesiredCapabilities

	desired_capabilities = DesiredCapabilities.PHANTOMJS.copy()
	desired_capabilities['phantomjs.page.customHeaders.User-Agent'] = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_1) ' \
	'AppleWebKit/537.36 (KHTML, like Gecko) ' \
	'Chrome/39.0.2171.95 Safari/537.36'
	driver = webdriver.PhantomJS(desired_capabilities=desired_capabilities)

	#!/bin/bash

	# Anh Nguyen <[email protected]>
	# 2016-04-30
	# MIT License

	# This script takes in same-size images from a folder and make a crossfade video from the images using ffmpeg.
	# Make sure you have ffmpeg installed before running.

	# The output command looks something like the below, but for as many images as you have in the folder.

	<!-- SEO / Google -->
	<meta name="author" content="Author name here....">
	<meta name="description" content="Description text here.....">
	<link rel="canonical" href="URL here...">

	<!-- Social: Twitter -->
	<!-- After inserting META need to validate at https://dev.twitter.com/docs/cards/validation/validator -->
	<meta name="twitter:card" content="summary_large_image">
	<meta name="twitter:site" content="@8bit_code">
	<meta name="twitter:creator" content="8bit_code">

	/* NOTICE: THIS WAS MADE BACK IN 2017, OF COURSE IT'S NOT GOING TO WORK WELL NOW THAT TWITTER'S FUCKED THINGS UP */
	@namespace url(http://www.w3.org/1999/xhtml);

	@-moz-document domain("twitter.com") {
	[data-component-context="suggest_recap"],
	[data-component-context="suggest_who_to_follow"],
	[data-component-context="suggest_activity"],
	[data-component-context="suggest_activity_tweet"],
	[data-component-context="suggest_recycled_tweet_inline"],
	[data-component-context="suggest_recycled_tweet"]{

	from __future__ import print_function
	import httplib2
	import os

	from apiclient import discovery
	from oauth2client import client
	from oauth2client import tools
	from oauth2client.file import Storage

	import datetime

	/*
	Copy this into the console of any web page that is interactive and doesn't
	do hard reloads. You will hear your DOM changes as different pitches of
	audio.

	I have found this interesting for debugging, but also fun to hear web pages
	render like UIs do in movies.
	*/

	const audioCtx = new (window.AudioContext \|\| window.webkitAudioContext)()