Zoom refers to the change of focus or scope of view from close distance to long distance or narrow to wide angle of view.  As far as surveillance cameras are concerned, there are 5 types of zoom specification that we have to be aware of for the selection of camera to achieve our intended purposes.

(a) Fixed focus lens with no zooming capability
(b) Variable focus lens for local adjustment
(c) Variable focus lens for local and remote adjustment
(d) Variable focus lens with criterion based control
(e) Digital zoom

Local adjustment in Type (b) refers to manual adjustment of the lens at where the camera is.  Remote adjustment in Type (c) refers to manual adjustment of the lens via a local area network or the Internet.  Type (d) is the subject matter of this article.

All the above 3 types are in the category of optical zoom.  Optical zoom abides by the law of nature as in physics.  Digital zoom is electronic or computer enabled.   Digital zoom can be applied to cameras with or without optical zoom capability.   It is an expression of the relationship between captured images in computer memory and displays on LCD screens for human eyes. 

Let us take the example of a 4MP (megapixel) camera capable of capturing images that are composed of 2032 x 1920 pixels or dots.  For static photography, this level of resolution is average.  For video surveillance, this is large as we rely on a LCD screen to view videos and we normally allocate a small window for one camera so that we can view as many as 16, 32 or 64 cameras at the same time.  Assume we have 12 windows on a 23” LCD screen and that the screen has 1920 x 1080 dots of maximum resolution.  Each window would have 480 x 360 dots only.  When we click on a particular window to enlarge its size, the bigger window will display more dots from the computer memory than the smaller window.  This is digital zoom.  Digital zoom will work until the window size reaches the captured image size.

Auto Zoom in this article refers to Optical and not Digital.  It is easy to program Digital Zoom to occur based on certain algorithms and this would be considered as auto zoom too.  Digital Zoom plays on images that have been captured.  Optical Zoom decides the level of resolution to capture.  For example we can set the camera on a wide angle for state of affairs monitoring and use 4MP of camera lens resolution for an area far away such as a horizontal width of 400m.  Each metre of the horizon is represented by 5 dots.  When a particular pre-defined spot is found to have motion detected, we can preset the camera to zoom in so that each metre of the horizon is represented by 100 dots for example.  This gives 20 times higher clarity.  All figures here are not mathematically correct and are shown for illustration of principle only.  The 3 slides below are sourced from ACTI and they illustrate the effects of Auto Zoom very well.

When night falls and the illumination level of the environment is very low, cameras normally automatically switch themselves to night mode and start to record images in black and white only.  Camera aperture would have changed and this means lens focus too.  Auto Zoom allows the camera to adjust its focus under this condition.  The 3 slides below are sourced from ACTI and they illustrate the effects of Auto Zoom very well.

Compucon K5211E camera has Auto Zoom capabilities of up to 18 times.

Zoom refers to the change of focus or scope of view from close distance to long distance or narrow to wide angle of view.  As far as surveillance cameras are concerned, there are 5 types of zoom specification that we have to be aware of for the selection of camera to achieve our intended purposes.

(a) Fixed focus lens with no zooming capability
(b) Variable focus lens for local adjustment
(c) Variable focus lens for local and remote adjustment
(d) Variable focus lens with criterion based control
(e) Digital zoom

Local adjustment in Type (b) refers to manual adjustment of the lens at where the camera is.  Remote adjustment in Type (c) refers to manual adjustment of the lens via a local area network or the Internet.  Type (d) is the subject matter of this article.

All the above 3 types are in the category of optical zoom.  Optical zoom abides by the law of nature as in physics.  Digital zoom is electronic or computer enabled.   Digital zoom can be applied to cameras with or without optical zoom capability.   It is an expression of the relationship between captured images in computer memory and displays on LCD screens for human eyes. 

Let us take the example of a 4MP (megapixel) camera capable of capturing images that are composed of 2032 x 1920 pixels or dots.  For static photography, this level of resolution is average.  For video surveillance, this is large as we rely on a LCD screen to view videos and we normally allocate a small window for one camera so that we can view as many as 16, 32 or 64 cameras at the same time.  Assume we have 12 windows on a 23” LCD screen and that the screen has 1920 x 1080 dots of maximum resolution.  Each window would have 480 x 360 dots only.  When we click on a particular window to enlarge its size, the bigger window will display more dots from the computer memory than the smaller window.  This is digital zoom.  Digital zoom will work until the window size reaches the captured image size.

Auto Zoom in this article refers to Optical and not Digital.  It is easy to program Digital Zoom to occur based on certain algorithms and this would be considered as auto zoom too.  Digital Zoom plays on images that have been captured.  Optical Zoom decides the level of resolution to capture.  For example we can set the camera on a wide angle for state of affairs monitoring and use 4MP of camera lens resolution for an area far away such as a horizontal width of 400m.  Each metre of the horizon is represented by 5 dots.  When a particular pre-defined spot is found to have motion detected, we can preset the camera to zoom in so that each metre of the horizon is represented by 100 dots for example.  This gives 20 times higher clarity.  All figures here are not mathematically correct and are shown for illustration of principle only.  The 3 slides below are sourced from ACTI and they illustrate the effects of Auto Zoom very well.

When night falls and the illumination level of the environment is very low, cameras normally automatically switch themselves to night mode and start to record images in black and white only.  Camera aperture would have changed and this means lens focus too.  Auto Zoom allows the camera to adjust its focus under this condition.  The 3 slides below are sourced from ACTI and they illustrate the effects of Auto Zoom very well.

Compucon K5211E camera has Auto Zoom capabilities of up to 18 times.

Compucon CPD Seminar
22 August 2012 Wed
Compucon House Albany

4:00 - 4:15pm Outside the Square
4:15 - 5:00pm Optimised Rack Systems Overview (TN)
5:00 - 6:00pm Heavy Data Systems Overview (Edmond)
6:00 - 7:30pm Wine and Cheese

Outside the Square

o We will talk about smart dust this time.  Smart dust has been deployed in war zones but not publicly announced for civil use as of today due to privacy issues.  We will look at it in more details than just saying it is the Internet of All Things to come.  We will cover the technology aspects of it and we will be surprised to realise how much we know (very simple indeed).  This will allow us to prepare ourselves for the near term future.

Optimised Rack Systems (ORS)

o This session takes a hardware platform perspective whereas the next session takes an implementation perspective.  ORS refers to the approach of maximizing installed performance capacity per unit of space or per watt of electricity consumed or per dollar invested.  Blade servers are the most known implementation of ORS, and container systems are the most recent developments of ORS.  These arrangements provide the context but they are not in demand by our average peers in New Zealand.  We will step down the ladder and introduce smaller implementations of ORS including the smallest one which is 1U2 meaning 2 systems in the space of 1U.  The main focus is on the FFP and TCO performance of these arrangements so that our peers can add this knowledge to solve computational problems or enhance organisational productivity and competitiveness.  We will compare them to non-optimized arrangements to assure that we do not lose any perspective.

Abbreviations: FFP is fitness for purpose.  TCO is total cost of ownership. 1U is 1.75” (44.45mm) of height in a standard 19” cabinet providing 42U of space.

Heavy Data Systems

o The industry uses the term Big Data to describe the very large volume of data for management and operational analysis.  We use the term Heavy Data System to describe the computing systems for handling big data.  Heavy is a term that encompasses a big range of weights.  We will focus on the lower half of the weight range in this session.  We will start from an 8 HDD DAS system and expand it to 800 HDD still as DAS (direct attached storage) and obviously with RAID protection and speed consideration.  Where does SSD fit in?  We will touch on server virtualisation as implemented by VMware and Microsoft.  We will talk about NAS and SAN as implemented by Microsoft and the Linux communities.  We will limit the scope to homogeneous computing only which is the norm for daily business requirements such as in a local area network of a small to large organisation.  We will talk about heterogeneous computing or super-computing in September as for research, design simulation, or digital content creation uses. 

Abbreviations: HDD is hard disk drive.  SSD is solid state disk.  RAID is redundant array of inexpensive disks.  DAS is direct attached storage.  NAS is network attached storage.  SAN is storage area network.