computer vision has been used successfully by many industries

let’s think of the applications for intelligent computer vision

you might work for the new york times

and use machine learning to digitize your best photo collection

secure restoring all your images

and finding new insights from the data locked on the backs of physical images

is only part of the task one of the biggest challenges in scanning a photo archive

is adding data regarding the contents of the images

state-of-the-art computer vision algorithms help you complete that task

or you might work for bugs and want to help your customers get more value from the images they store in your platform

you might want to make image files as easy to find

and search through as text documents

to do that you need the technology to provide high quality image labeling

and scale to the massive number of image files stored in your repository

here computer vision helps you classify images into categories of similar images

analyze the content based on user requests

and it also returns the results

and a score of confidence in the analysis.

The ability to automatically classify and label images is a powerful tool for bugs customers

or you might be a researcher in hart research institute at texas a m university corpus christi and want to identify

the types of shorelines between aerial imagery of the cost in order to accurately predict the environmental sensitivity index esi of shorelines displayed in images

this index classifies the sensitivity of a section of shoreline to an oil spill

computer vision can be used to accurately classify different types of shorelines in aerial images

computer vision as a field

is more than just a binary image classification tool

Vision API is a pre-built ML model created by google

that lets you pass through a json request and receive a ranked list of associated labels for an image

and if a photo contains

more than one object

the api can draw bounding boxes and also classify pieces of an image

modern image classification models can even generate captions that describe the action in the image, for example, it can return a result from a picture such as

two hokey players are fighting over a pug

it’s important to note that even the best models can make mistakes in their label assignments or predictions like in this example of a road sign captioned as

a refrigerator filled with lots of food and drinks

machine learning has many impressive uses

such as detecting objects

texts

faces

or poses in images

helping detect disease

and even enabling cars to drive themselves

the key point is that computer vision models

can automate tasks that might assist a human team in fact the latest computer vision models perform better than humans in some tasks.