What is Computer Vision? Top 10 most effective business use cases

Welcome to this executive briefing on Computer Vision.

Your time is valuable, and with that in mind, this article will provide you with only the most critical information necessary to quickly understand Computer Vision and see concrete examples of how it can greatly benefit your business.

First, we will talk about what computer vision is and why it has such a high potential to disrupt virtually any business and any industry. Then we will discuss how computer vision works and you will find out about two types of computer vision.

Then we will dissect an actual example of how computer vision has disrupted an industry. We will look at a fruit and vegetable sorting example.

Well, this might be a very trivial use case. It might not perfectly be representative of your industry.

Then we will dive into what possible applications of computer visions exist out there to help you come up with ideas of how this exponential technology could potentially be applied in your industry and in your business.

And finally, we will look at ten different case studies of computer vision already being applied in different companies and industries.

Table of Contents

What is Computer Vision?

Computer vision is a field of artificial intelligence that trains computers to interpret and understand the visual world.

Computer vision has been around since the 1950s.

For example, the barcode scanner is an example of computer vision.

However, only in the past decade, thanks to exponential advancements in research and technology. Computer vision has become so widespread, computer vision is permeating our lives.

In fact, we often take it for granted, here are few applications from our daily lives:

  • When your camera identifies a face and our focus is on it, that’s computer vision in action
  • When Facebook suggests whom to tag in a photo, that’s also computer vision in action
  • When a car can drive itself, that’s also computer vision

Computer vision has huge potential in business solutions. It is one of the easiest technologies to implement in business.

The reason is that we humans process about 90 percent of data in visual form, meaning that previously a huge variety of tasks relied on human vision from traffic inspectors to MRI scan analysis experts from sport umpires to check out assistance from truck drivers to equipment inspections.

The list goes on with the power of computer vision. All of these tasks can now be automated. Computers will do them faster and with higher accuracy. Businesses can increase efficiency, save costs, and human workers can be freed up to do more creative tasks.

How does Computer Vision work?

Computer vision works only through deep learning, i.e. through neural networks.

You will hear this all around. The other day I even watched a video published by someone from Google titled How Computer Vision Works, which assumed that this was correct.

This is only half of the story. There are two types of computer vision methods:

  • Classical Learning  
  • Deep Learning.

Classical learning

It relies on pre-built libraries of features.

For instance, if you’re trying to detect a face, these libraries of features will allow your algorithm to find the eyes, the nose, mouth and so on very quickly.

There are libraries for most common objects, from knives and forks to cars. There are no neural networks involved, and the beauty of classical computer vision is that it’s efficient and often gets the job done extremely fast.

Deep learning

The types of neural networks that make this possible are called convolutional neural networks.

The best way to think about it is that we have a black box inside which there’s a neural network and indeed it’s like a black box because once we let it learn, we don’t control how it learns. There is no pre-programming.

But all we know is that this black box has a capacity for memory or experience. What we will do is feed this black box tens of thousands of images of, let’s say, dogs and cats.

All these images are labelled. And over time, the neural network will have seen enough images of dogs and cats to automatically extract features that it thinks are distinguishing for each of the two animals.

For example

Cats have pointy ears, dogs have fluffy ears, cats have whiskers, dogs have big noses.

In this way, the neural network will create features for its own future use.

Then when we feed it a previously unseen image of a dog or a cat, it’ll be able to apply those features to understand what animal it is processing. This is a very trivial example, but in a similar way, convolutional neural networks can be taught to identify any type of object, for example, recyclable plastic bottles versus non-recyclable rubbish as long as there’s ample label data available.

Tomato Sorting Example:

Before we list various applications of computer vision, let’s look at one specific example, which will illustrate how easy it actually is to innovate with computer vision.

The example we’re going to look at is tomato’s sorting.

Let’s say that you are the CEO of a company that wholesales tomatoes to grocery stores.

You want to make sure that your produce is of highest quality and therefore on your conveyor belt, you have three people looking at the tomatoes and removing the unripe and rotten ones.

Drawbacks of this approach are as follows:

  • First, you need to pay three people to do this work all the time
  • Next is that humans make mistakes. Therefore, quality is not ideal drawback
  • Humans are slow, slower than technology drawback
  • Humans might get sick and not come to work, and this can cause the whole process to break down
  • This is not really enjoyable or creative work that these people really deserve to be doing

So now let’s see how we could innovate with computer vision.

What do we do is we would show a computer vision algorithm, lots of labelled examples of good tomatoes and lots of examples of unripe and rotten tomatoes.

In this case, you would be using deep learning type of computer vision algorithm.

The good news is that you already have the labeled data. It’s part of your process. You just need to set up a camera above the workers and lets the algorithm watch their actions.

It will see the tomatoes that go through, and those images will be classified as good to go, and it will also see the tomatoes that get discarded and those images will be classified as unripe, rotten.

After a week or so of observations, the algorithm will have seen enough training data and from then onwards it can make the same decisions autonomously.

All that’s left to do is add a simple robotic mechanism that will knock out the rodent or unripe tomatoes.

The advantages of this approach are as follows.

  • Costs are minimized
  • Errors are minimized
  • Speed is maximized
  • Humans can do more meaningful, creative and enjoyable work

Applications of Computer Vision

In this section, we will look at various applications of computer vision, all of these can be applied to images as well as videos, because videos are simply many frames following one another.

Image classification

  • It groups images into different categories and answers questions such as,
  • Is this an image of a dog or a cat?
  • Is this an image of a defective or a working part?

Image segmentation

  • It partitions an image into multiple regions or pieces to be examined separately.
  • How Skype can blur out the background behind you on a call application

Object detection

  • Is there a pedestrian in this image or not?
  • Is there a water spill on the grocery shop floor or not?
  • Advanced object detection can recognize multiple objects in one image and draw boxes around them
  • Computers have surpassed humans in specific applications of object detection

Object tracking

  • Where is this pedestrian going?
  • Is that car traveling too fast?
  • When a Tesla car was able to apply brakes on a highway pre-emptively, meaning the brakes were actually applied split seconds before and it collision happened in the same lane ahead of that vehicle.

Image generation

  • You could transfer the style of wonderous onto another
  • You could generate a 3-D representation from 2D images. That’s exactly how Google Maps does it.
  • When you go into the mode when you can see three dimensional images

Edge detection

It identifies outside edges of an object or landscape to better identify what is in the image application.

Face detection

  • What should the camera focus on?
  • How many faces are in this image? That’s how our phones can detect where our faces are in photos

Facial recognition

  • This is a next step up and beyond facial detection and can identify specific individuals who should be tagged in this photo

Optical character recognition, or OCR

  • It recognizes words and numbers in scans or even handwritten documents
  • Computers have gotten very good at this, achieving accuracy rates of over ninety nine percent

Pattern detection

It recognizes repeated shapes, colors and other visual indicators and images.

Feature matching

It is a type of pattern detection that matches similarities in images to help group them together.

10 Use-cases – with Real Results

Skanska company

Computer vision actually adding substantial value to companies already.

Uses companies in the world, determined that their workers walked

an average of about six miles per day to get the right materials, tools, and equipment to the right place at the right time, using computer vision to track their movement and compare it to their activity data.

  • The company was able to optimize the entire building process, positioning workers tools and resources where and when they needed to be
  • This saved each worker about 2 miles of walking every day, which in turn boosted the productivity of each worker by an hour, every single day, roughly about a 12 % increase


Uses computer vision in their self-driving cars.

If you want to drive a car, you need to be able to sense your surroundings, and while hearing plays a big part, the most important sense in driving is sight.

Similarly, with self-driving cars use a multitude of scanners to analyse their surroundings. But computer vision is crucial in creating a safe, comfortable driving experience.

  • Since 2016, all Tesla cars have full self-driving hardware, which includes 8 cameras that allow the car to see its environment at centimetre precision
  • This move coincides with Tesla’s massive growth in revenue, which grew by around 400x and 50% in the 3 years

Harvest S.R.O. Robotics

Uses computer vision to enable automatic crop harvesting.

  • Even though we think of harvesting as a process done by huge machines, a lot of it is still done by hand
  • In Florida alone, 10 to 11 thousand acres of strawberries are picked by hand
  • Every single year is predicted that the number of agriculture workers will reduce by six percent by 24
  • Harvest S.R.O. Robotics has developed a robot that uses computer vision to inspect strawberries, determine which ones are ripe and pick them all in eight seconds per plant
  • One of these can harvest 8 acres per day, the equivalent of 30 human workers


Uses computer vision for their image search capabilities.

Consumers worldwide purchase around 3 trillion dollars on the Web every single year, with the fifteen to twenty five percent growth predicted to continue into the foreseeable future.

  • Online retailers are coming up with new ways to improve the shopping experience and increase their revenue
  • A recent innovation is image search. If you see something you like, you can take a photo of it and uploaded to find it on eBay
  • The program uses computer vision to analyze a picture and search its entire inventory of over a hundred million items to find what you’re looking for
  • Is it difficult to estimate the direct impact of this on eBay’s revenue? But it’s telling that half of all sales now are done via mobile, compared to only 40 percent back in 2016

Medication adherence

  • Adherence to prescribed medication is crucial for successful recovery, but it is quite low due to its nature
  • Adherence is difficult to measure, but it’s estimated to be around 50 percent, earning that approximately half of all patients don’t take their medicine as prescribed
  • A cure developed smartphone software that monitors patients using computer vision to determine whether they took the right medication at the right time and in the right amount
  • They achieved a 95% adherence on initial samples and could significantly help most vulnerable groups such as substance abusers and people with mental disabilities use


Uses computer vision for automatic oil well monitoring

  • Monitoring large numbers of production facilities is costly and time consuming
  • Osprey informatics employed computer vision algorithms to monitor remote oil wells
  • This allowed them to eliminate unnecessary visits to a functioning site and react to malfunctions quicker
  • The costs of monitoring were reduced by half, and oil giant Shell quickly invested several million dollars into the state

Waste storage

We generate over two billion tons of waste every year, and although around seventy five percent of it is recyclable, we only recycle around 30 percent of it.

Even the top countries in the world, like Switzerland, Australia and Germany only recycle about half of their waste. This isn’t just bad for the environment.

  • It’s a huge waste of money with 11 billion dollars’ worth of recyclable packaging being thrown away every year in the US alone
  • Roaf, the world’s first fully automated sorting plant for municipal waste, opened in Norway in twenty sixteen, using computer vision to perform soring work that used to be done manually by workers in 12 hour shifts
  • The sorting process has a capacity of forty tons per hour, serves one hundred and ninety thousand inhabitants and is up to ninety seven percent accurate in sorting use

Swiss Federal Institute of Technology

Uses computer vision for stroke recovery every year

  • 15 million people across the world suffer stroke, and although a majority survive, stroke remains the leading cause of serious long term disability in developed countries
  • Recovery is with dilatation services during the first year costing an average of 11 thousand dollars in the US, many are unable to afford this and they often suffer permanent damage that could have been avoided, reducing their quality of life and their contribution to society
  • Swiss researchers have developed a computer vision program that monitors coordination, dexterity and reflexes recovery as a result of brain stimulation, something that originally needed to be measured by trained, expensive professionals
  • While the technology is still in early stages, initial testing has shown full recovery effectiveness and eventual implementation will change the lives of millions of people

Workplace safety monitoring

Businesses that deal with industrial machinery such as construction or manufacturing, deal with numerous workplace safety hazards, and average business in these fields will experience 27 injuries that lead to days off every year, almost half of which will be due to human error

  • An astonishing 30 % of all injuries are simply due to workers not wearing the proper protective gear at the right time
  • Texaco’s software allows businesses to monitor the workplace 24/7 and issue alerts
  • When a worker is not wearing the gear, they should be. They predict that the full adoption of this system could prevent sixteen thousand accidents, costing a total of one hundred and 22 million pounds every year in the UK alone

Tamira Automatic or sorting

The mining industry consists of many processes with high energy requirements and consumes around 3% of the world’s energy. This high energy use is largely due to precious materials being present in or in very small quantities

  • Average gold bearing contains about 5 grams of gold per ton
  • Tamira developed a range of automatic sorting machines that use computer vision to identify particles with larger volumes of valuable materials and separate them from the waste using short bursts of air
  • This system can reduce the amount of required energy by 15% and the amount of water by up to 4 cubic meters per ton

So, there are the 10 effective use cases for computer vision and the list doesn’t stops here, as there are many more and growing but you got the idea.


So in a nutshell,

Computer vision can help you automate any kind of tasks that rely on human vision.

The processing of visual data, which is 90% of all data, can be automated with computer vision, leading to increased efficiency, cost savings and freeing up human workers for more creative tasks.

Looking ahead, the future of computer vision holds tremendous promise and offers limitless opportunities. As technology continues to advance and the data available for machine learning increases, the potential applications of computer vision will continue to grow.

In industries such as healthcare, finance, transportation, and more, computer vision will play an increasingly important role in automating tasks and improving decision-making. The future of computer vision is exciting and full of possibilities, and it will be fascinating to see how it continues to shape our lives and change the world around us.

So, there we go.

I really hope that now you are much more comfortable and confident speaking about computer vision and more importantly, seeing how you can apply them in your industry and in your business.

Thank you for reading our blog, we hope you found the information provided helpful and informative. We invite you to follow and share this blog with your colleagues and friends if you found it useful.

Share your thoughts and ideas in the comments below. To get in touch with us, please send an email to dataspaceconsulting@gmail.com or contactus@dataspacein.com.

You can also visit our website – DataspaceAI

Leave a Reply