Wednesday, April 27, 2016

Limitations of Computer Visions

Computer vision can be defined as a professional facility that boost a computer to notice or understand the content of a picture. It is further classify as one of the major frontier of technology in computer science which helps society to facilitate their day to day activities.

             Of course, computer vision is of a great benefit to individuals but it it has its own demerit. An expert scientist who is involved in the development of artificial intelligence(AI) has warned that it could be able to rise up against humans within decades but if managed correctly it could bring great changes to society.
            As for the demerits, computer vision tool-set are rare and too costly however, it depends on the quality of the image captured so it needs a high definition camera for each and every project to produce a perfect result. Besides, It needs special maintenance to perform effectively and consistently.
            Moreover, computer vision shows a vivid difference between people. In the sense that it create vast job opportunity for the highly intelligent ones and make the average or poor masses lose their job for an immortal robot with no family to support neither wife or children to take care off. In addition, with the rapid growth of technology one will wish that scientist should come up with different ideas on how different categories of people will live peacefully without these demerits affecting them.
Therefore, advantages and disadvantages work with the principle of direct proportionality. Which means without advantage there will be no demerits and scientist love challenges so there will always be room for disadvantages.
Reference
http.www.google.com/image search

www.intechopen.com/books/computer vision

Programming Languages Used In Computer Vision

Few decades ago, It was almost impossible finding a decent library for computer vision. Computer vision engineers had to use the oldschool way picking up a pen and a paper and had to write down their own library of computer vision algorithm.

Now that many libraries are available, The question that made people have mini debates was, Which program 

Monday, April 11, 2016

Brief History of Computer Vision

Source: Google

As any other technology the world has ever witnessed, Computer Vision did not come to us as a revelation just over night. What we see or know about Computer Vision and it's related fields are results of several years of  research and hard work. The first such works started in the 1960s. By then, light capturing digital hardware has already been invented paving the way for scientists and engineers to tackle the challenge of Computer Vision.

1964

Defense contractors Woody Bledsoe, Helen Chan Wolf and Charles Bison launch a facial recognition system for an unnamed intelligence agency. Further developments and iterations went on to change how governments and security bodies such as police match and identify citizens and also suspects from ID photographs stored in their national databases.

1966

Minsky assigns computer vision as an undergrad summer project. The first attempt to solve this problem was made by Seymour Paper. Though this attempt was not as successful as intended, this got scientists and engineers talking.

1970's

This decade saw some progress on interpreting selected images. One advancement was UK Police inventing a license-plate recognition system in 1976 which was deployed around London in 1993 to counteract the threat of IRA bombings.

1980's

Researches attempt using Artificial Neural Networks in Computer Vision to allow for better image recognition but however they quickly abandoned the study as it deemed too difficult with the limited resources of the time. There was a shift toward geometry and mathematica rigor. This decade saw one of the very first beginnings of Vehicle Autonomy using Computer Vision developed by Lockheed Martin, Carnegie Mellon and others. It is a land vehicle that uses video based imaging to follow a road at just 3 mph.

1990's

This decade can be said as the actual birth of Computer Vision that saw the return of Artificial Neural Networks in Computer Vision research under the name, Convolutional Neural Networks and were able to solve a challenging for the time problem, that is digit recognition for bank cheques. It also saw further advancements in Facial Recognition and statistical analysis in vogue.

2000's

Source: Google Images
Reached new heights in object recognition due to the availability of large annotated datasets and video processing began. In 2004, Mars rovers Spirit and Opportunity uses Computer Vision to land successfully in Mars. The first 3D pizza system called called the "Scorpion" builds a 3D profile of 7,200 products per hour using multiple cameras.

2010's

Source: Google Images
Computer Vision began to be deployed in a number of web applications and mobile apps and saw these technologies reach a wider user base. Microsoft Releases Kinect in 2010, a motion sensing camera that can track 20 human features at 30 times per second. Google starts testing it's autonomous cars which uses cameras, LIDAR and artificial intelligence to navigate unassisted in 2012. In 2014, smartphone processors become powerful enough for pattern recognition. Many mobile photography applications could automatically enhance photos and suggest best taken photos from a session of photo shoot.

Today, computer vision is rapidly being utilised in almost every field from, social to health applications, entertainment industry and also military drones and further research is being made in the field. What we are witnessing is the Golden Age of Computer Vision.

References:
https://tordivelblog.com/2014/06/23/scorpion-3d-pizza-sorter-57-years-of-machine-vision-history/
https://cs.brown.edu/courses/cs143/lectures/01.pdf
http://www.egavves.com/category/home/#sthash.PtZp3lvq.dpbs
http://www.consciousentities.com/minsky.htm


Artificial intelligence (AI)

Image result for artificial intelligence
Source: Google Images    


(AI) is the intelligence exhibited by machines or software. Computer Vision mostly is associated with or part of artificial intelligence. It is also the name of the academic field of study which studies how to create computers and computer software that are capable of intelligent behaviour. (Wikipedia) 

It can also be defined as the  science and engineering of making intelligent machines, specifically intelligent computer programs. It's  similar to using computers to understand the human intelligence. However, AI does not have to confine itself to methods that can be observed biologically.


                                                   Source: Google Images      
Can We Predict The Future of AI?  

Raymond Kurzweil, A very well know American author, inventor and futurist has made many future predictions regarding Technology, The Internet and their explosive growth and how it is going to impact our lives. 

In  The Age of Intelligent Machines, is one of his books that was published in the early 90's Where he predicted that the Internet would have a marvellous impact   in the number of users as well as the content, It grants the users access "to international networks of libraries, data bases, and information services. 

Kurzweil correctly foresaw that the preferred mode of Internet access would inevitably be through wireless systems, and he was also correct to estimate that the latter would become practical for widespread use in the early 21st century. 


Image result for chess game man vs computer
He also  extrapolated the performance of chess software to predict that computers would beat the best human players "by the year 2000"  In May 1997 chess World Champion Garry Kasparov  was defeated by IBM's Deep Blue computer in a well-publicized chess tournament.

                                             

                                                                                                                   Source: Google Images



                                     Humans Vs Technology                                 
One of the predictions that made many people afraid of the future, Is that by the year 2030 a robot can be smarter than all mankind combined. Technology has already replaced mankind in many industrial fields which lead to a huge decrease in the global labour force. (after all they won't need to pay the robot a salary). If robots will think on our behalf one day, That can only mean that the human intelligence would be decreasing. 

We used to wash our clothes by hands and then replaced it with washing machines. we also used Calculators as a replacement of our brains to compute simple maths. As well as machines that replaced the labour force, And many more example that we can see in our daily life.

I can only hope that the Artificial intelligence is going to help make the world a better place. And work on the service of humanity. We can't be civilized if everything is improving and humans are doing the opposite.



Image result for robot serving humans                                                             Image result for terminator                                      Source: Google Images 

      

References: 
https://en.wikipedia.org/wiki/Artificial_intelligence

https://en.wikipedia.org/wiki/Predictions_made_by_Ray_Kurzweil

http://www-formal.stanford.edu/jmc/whatisai/node1.html