DRI in Video Surveillance

DRI in Video Surveillance

Wish you a very happy new year 2018. In today’s competitive environment, having new customers constantly is one of the keys of success. To make professional quotes with proper sketches, realistic test objects, 3D images, mock ups etc, which leave a lasting impression on your clients.

When you are answering queries related to your product, you can quickly glance through the details while your client is on call.

DRI stands for Detection, Recognition and Identification in video surveillance. DRI ranges, expressed in meters, km (or miles), can be found in the specification table of infrared camera brochures. In order to select the right sensor meeting the application requirements, these DRI ranges have to be, first, perfectly defined, but also assessed with regards to globally adopted industrial standards.

DRI Definition

The terms “Detection”, “Recognition” and “Identification” were defined as follow:

• Detection: ability to distinguish an object from the background
• Recognition: ability to classify the object class (animal, human, vehicle, boat …)
• Identification: ability to describe the object in details (a man with a hat, a deer, a Jeep …)

The following pictures illustrate these definitions:

    Left image: Detection – At several kms, 2 targets are detected out of the background 

    Center image: Recognition - a human is walking along the fence 

    Right image: Identification – 2 males with trousers and jackets are identified – one is smoking.

We detect an object when it enters the field of view. Detection means we are aware that an object (or person) now exists where previously it was not seen. Usually, this is due to movement of the object into the field of view of the surveillance camera. Detection simply means we are aware of the object, but have too little details to recognize or identify the object.

As the object moves closer, we may recognize the object from characteristics previously encountered. For example, aircraft recognition is taught to military ground troops and airmen. All aircraft have wings, engines, a fuselage, and tail assembly. They differ in size, shape, number, and position to each other. A particular model of aircraft can be recognized by recalling these characteristics from pictures, drawings or past detailed observations.

Identification is the process where sufficient details are available to uniquely discern a person or object that is previously unknown. Identification requires sufficient detail to accurately describe or recall the characteristics of the subject at a later time. For example, a mug shot (booking photograph) is taken following the arrest of a subject as a means of photographing (recording) sufficient details for later identification by a victim or witness. In video surveillance terms, sufficient detail is calibrated in pixels per foot of the area recorded by the camera.

DRI Ranges

“A picture is worth a thousand words” goes the old and equally wise saying. It is true that most human beings learn much quicker when aided by visuals rather than pure text as that is more in tune with the human psychology of learning.

So instead of remembering commands and functions you will easily remember that yellow color indicates areas where it is possible to recognize people, red color indicates areas suitable for people identification, green areas for detection and so forth.

To put this in perspective, sensors have a resolution of 640×480 which is over 300,000 pixels.
Human “detection” only requires 3.6 of those pixels and “identification” only requires 230 pixels, which is an extraordinarily small amount on the screen that can easily go unnoticed by the human eye. In fact, if this page were the size of your video feed, the area required for a human detection rating is about equivalent to a lowercase letter “i” in this text. The amount of detail visible at the detection, recognition and identification distances is not as high as one might expect, as can be seen in the chart below.

Another thing that is often not mentioned is that these ratings are based on what is termed “ideal conditions” which rarely happen in the real world. The average environmental application will get 25% less than the distance that the thermal camera is rated for and in extreme conditions can be less than 10% of the rated distance.

Based on SR-100 & SR-100P by FLIR the approximate DRI (Detection, Recognition and Identification) range for a vehicle and a human target is listed in Tables 1 & 2.

Lens	Detection	Recognition	Identification
100mm	4.4km	1.1km	580m

Table 1: DRI range for Vehicle with 2.3m critical dimension using a Thermal Security Camera with 38 micron pitch detector

Lens	Detection	Recognition	Identification
100mm	1.6km	400m	200m

Table 2: DRI range for Human Target using a Thermal Security Camera with 38 micron pitch detector

Assumptions:  50% probability of achieving objective at the specified distance given a   2-degree temperature difference and 0.85/km atmospheric attenuation factor.  Actual range may vary depending on camera setup, user experience, environmental conditions, and type of monitor or display used.

Each & every Camera has this DRI ranges, before preparing project oriented ppt, do confirm the ranges & show in a slide with sketch.

Artical publish at safe secure magazine January 2018 issue.

Warrior 2.0 26BF2 is a 1080P 2MP Fixed Lens Mini Bullet Camera

Detect: 100 ft
Recognize: 50 ft
Identify: 25 ft

Laser 2.0 - 26ZV-LIR is a 2MP (1080P) IP PTZ Camera with 33x Optical Zoom & Laser Infrared camera and Spotlight 2.0 - 26ZV-W is a 2MP (1080P) IP PTZ Camera with 22x Optical Zoom & Full Spectrum Light Camera

Detect: 950 ft
Recognize: 875 ft
Identify: 750 ft

Lookout 2.0 - 26ZV is a 2MP (1080P) IP PTZ Camera with 30x Optical Zoom Camera

Detect: 1000 ft
Recognize: 900 ft
Identify: 800 ft

Judge 8.0 - 26DV8 is a 4K (8MP - 4x1080P) Multi-Purpose Lens Dome Camera with Motorized Zoom and IK10 Camera and Gladiator 8.0 - 26BV8 is a 4K (8MP - 4x1080P) Multi-Purpose Lens Bullet Camera with Motorized Zoom and P-Iris Camera

Detect: up to 200 ft
Recognize: up to 150 ft
Identify: up to 100 ft

Deputy 4.0 - 26DF4 - 4MP (2x1080P) Fixed Wide Angle Lens Turret Dome Camera

Detect: 115 ft
Recognize: 50 ft
Identify: 35 ft

Archer 2.0 - 26BV2-L - 2MP Long Range, Low Light Bullet Camera with Motorized Zoom and Focus

Detect: up to 450 ft
Recognize: up to 350 ft
Identify: up to 300 ft

Warrior 4.0 - 26BF4 is a 4MP (2x1080P) Fixed Lens Mini Bullet Camera

Detect: 150 ft

Recognize: 75 ft
Identify: 50 ft

CCTV Recording Resolution

Digital Video Recorders (DVR) and Network Video Recorders (NVR) are the heart of every security camera system. Customers often ask us about recording quality from CCTV cameras, things can seem relatively straightforward: you’ll need a DVR/NVR that can handle as many channels as you have cameras, and the more hard drive space you have for archiving all that video, the better. Naturally a customer wants to know how clear the video resolution will be from a system before making a purchase decision. Final outcome of your captured footage, 

Recording resolution is the number of pixels (dots) used to create an image. Higher resolutions use more pixels to create an image. This means that greater amounts of detail can be expressed in the image, but larger files sizes and a greater amount of storage (i.e. Hard drive space) are required to save the images or video. See the example below.

The resolution is increasing from left to right

Recording resolution is measured as the number of horizontal pixels by the number of vertical pixels (Width x Height). The following table shows some common recording resolutions.

Name	Width (Horizontal Pixels)	Height (Vertical Pixels)	Total Number of Pixels	Recommended Camera Equipment
CIF	360	240	86,400	320TVL, 400TVL, 420TVL, 480TVL,
2CIF	720	240	172,800	420TVL, 480TVL, 540TVL
VGA	640	480	307,200	540TVL, 600TVL, 1MP
D1	720	480	345,600	600TVL, 650TVL, 700TVL, 800TVL
720p	1,280	720	921,600	720p, 1.3MP
QVGA	1280	960	1228,800	960p, 1.3MP
1080p	1,920	1,080	2,073,600	1080p, 2.1MP
3MP	2048	1536	3145,728	3MP
5MP	2560	1920	4915,200	5MP
6MP	3032	2008	6088,256	6MP
8MP	3264	2448	7990,272	8MP

Notice the increase in the total number of pixels as the resolution increases. Because the total number of pixels is determined by the number of horizontal pixels times the number of vertical pixels, if both horizontal and vertical pixels are doubled, the total number of pixels increases by a factor of 4. This means that the amount of hard drive space needed to store an image or a given duration of video will also increase by a factor of 4 (given factors such as the frame rate remain equal). It is recommended to try different configurations on your system to balance image quality against the amount of storage space available.

Now based on the capabilities of your DVR you may choose to record in a certain picture resolution based on certain requirements unbeknownst to anyone but you.

I have seen a lot of misconceptions of the newly introduced 960H. Amongst the confusion, I have seen a statement along the lines of 960H provides you a 960x480 and/or 720x480 picture; as well that it improves the picture of any camera fitted with a 960H sensor via your current DVR. This is a BIG misconstrued myth. Submitted for your approval are the facts of 960H

1.   960H is NOT a megapixel resolution

2.   960H sizes are NOT 960x480 or 720 x 480 formats

3.   Cameras equipped with 960H alone, will NOT provide you with a refined captured footage of 960x480

The graphic below shows a comparison of the resolutions listed if each pixel takes up the same amount of space.

F.Y.I For all footage across every system, after video is recorded it can be magnified (zoomed in) but only digitally either through a computer or the DVR (if capable). That means, the individual pixels that create the picture can be made bigger. However, for analog footage, there will be no real advantage at detail due to the low-res of pixels.

Analog vs. Digital Resolution – TVL vs. Pixels

One of the most confusing and difficult topics in the CCTV world is resolution. Most of us have digital cameras or video camcorders and have heard the term megapixel used as the most common comparison in resolution between various makes and models. We are also aware that a larger number means better picture quality, but many people do not know why. In the CCTV security camera world, though, most cameras are still analog and their resolution is measured differently from what we are used to.

When measuring analog resolution, a TV line does not have a defined number of individual pixels. Instead, the term “TV lines” refers to the number of discernable horizontal or vertical lines on the screen. Analog security cameras are measured in Analog TV Lines, and most of them have between 420 and 580. The higher number of TV Lines, the more information captured. These types of cameras connect to a security DVR or CCTV VCR via coaxial video cable.

Zoomed too far into a picture from a website and seen the image go from clear to a bunch of colored squares – each one of those squares is an individual pixel. A megapixel (MP) is 1 million pixels, and is a specific measurement for digital resolution that encompasses the area of the output video.

Example: If a camera outputs a signal that is 1280×1024 pixels, it is shooting at a megapixel resolution of 1280 x 1024 = 1,310,720 pixels = 1.3 Megapixels (MP).

The most common type of digital security cameras are IP Cameras. These, like your digital camera at home, use strictly digital resolution. They utilize a network connection to either act as a standalone device or connect to a network-based DVR (Digital Video Recorder) or NVR (Network Video Recorder). IP Cameras have fixed resolutions and are now approaching, and in several cases exceeding, 1 megapixel in resolution, on average. Many of these cameras also support POE (Power Over Ethernet), which allows them to be powered by the Ethernet cable used for network connectivity, and PTZ (Pan Tilt Zoom), allowing for remote control of the pan, tilt and zoom features, if applicable.

I hope this helped everyone distinguish the differences between Analog (TV Line) and Digital (Pixel) Resolution. We would love to know what you think of our articles, and if you have any further questions, don’t hesitate to leave a comment!

RAW Formats

RAW Format implies that there is no compression done on the image. The major types of RAW format are RGB, YUV, YIQ. Our eye is more sensitive towards light intensity variation than color variation. So loss on color information will not affect the over all quality of the image. RGB is an end stream format. Information from the image sensor is in RGB format and we need the same format for displaying the image on an end

device. YUV & YIQ formats are developed for Analog TV transmission (NTSC & PAL respectively) and the digital version of YUV, YCbCr is the most common format used for image and video compressions.

Conversion from one format to another is described below:

RGB to YCbCr Conversion

Y = 0.299 R + 0.587 G + 0.11B

Cb = 0.564 (B - Y)

Cr = 0.713 (R - Y)

YCbCr to RGB

R = Y + 1.402 Cr

G = Y – 0.344 Cb – 0.714 C r

B = Y + 1.722 Cb

Y – Luminance Signal

Cb, Cr – Chrominance Signal, Color difference signal

R – Red

G – Green

B – Blue

Need for Compression

Consider an image of resolution 640 × 480. Let us calculate the size of the picture in RAW format. Each of the 10 Color is represented by 8 bits. Then for each pixel it needs 24 bits. Total no of pixels in the image is 640 × 480 = 307200 pixels. Therefore the size of the image turns to 307200 × 3 bytes = 921600 bytes. But an image in compressed format with the same resolution takes only 100 KB.

In the case of RAW video stream of length 1 sec its needs 640 × 480 × 3 × 25 = 23040000 bytes (23 MB) of storage if the frame rate is 25 frames/sec. But it’s known that the VCD format video having a size 700 MB plays for around 80 minutes. In the former case we need 110400 MBs (23 MB × 60 × 80) as storage space for 80 minutes video. Therefore we can achieve a high compression 150: 1 at the cost of computational complexities.