Network Camera 2014 Update: Standard Features and Functions

Over the past few years, network cameras have broken out of their shells and began to gain acceptance and adoption. Meanwhile, the network video surveillance market continues to be headed to a “tipping point” in 2014. This can surely be attributed to the maturity of the IP-based cameras and systems available today, as well as the increasing affordability due to intense competition from manufacturers all around the world.

According to a report from IMS Research, part of IHS, revenue for network video surveillance will overtake analog sales for the first time ever in 2014. Also, over 75 percent of new enterprise projects are employing IP cameras in their constructions. Not only that, but IP cameras are now increasingly being adopted by the SMB and residential sectors with the increasing popularity of cube ip cameras, since homeowners and small to mid-sized business owners can monitor their own stores without having to spend a fortune on surveillance solutions that are overly complex for their environments.

Based on another report by Allied Market Research, the combined North American market for IP-based video surveillance and video surveillance as a service (VSaaS) is forecasted to expand dramatically, rising from US$2.4 billion in 2013 to more than $18 billion by 2020. As these reports suggest, network video is and will be the mainstream for years to come.

With so many flashy features available, the following are some of the most important features or functions in a, or for a, standard network camera. After making massive improvements and advancements, what else is next? This year, it seems many manufacturers are leaning in favor of the low-light function.

LOW LIGHT ALLOWS VISIBLE COLOR IN NIGHTTIME ENVIRONMENTS

The low-light function is the ability of a surveillance camera to capture color images in environments with weak or no lighting. The performance of a low-light camera is highly dependent on the sensitivity of its sensor and the processing ability of its DSP, in combination with its iris type and shutter speed. The iris allows and controls the amount of light to pass through the lens and to the sensor inside the camera.

By working together, they are able to regulate the camera’s light input. New imaging technology as well as higher resolution lenses help to improve overall sensitivity, which continue to improve results from low-light cameras.

Typical day/night surveillance cameras automatically switch to black/white mode once it gets dark, or when lights are shut off. But low-light cameras are becoming increasingly popular as being able to tell colors apart at night is important for identifying suspects and culprits. Low-light cameras are also crucial to settings such as mechanical rooms or critical infrastructures, where the color of lights on different equipment must be discernible. For example, those monitoring in the control center must be able to tell the difference between blue and green lights or red, orange, and yellow lights in order to evaluate the situation in the mechanical rooms and critical infrastructures, and immediately tell if there is an emergency or breach. Presently, the most common lighting condition for a low-light camera to visibly display color images is at 0.01 lux. For more advanced cameras, they are able to visibly display color images in low-light environments up to 0.001 lux.

Now, security manufacturers and their R&D teams are aiming to clearly differentiate colors in nighttime environments up to 0.0001 lux, which is quite a challenge, as it is difficult even for human eyes to see color in this condition. Furthermore, there is no standard test for manufacturers to test their cameras against at the moment. Each manufacturer has its own different set of testing standards which causes inconsistency in camera performance with each different brand. Also, certain testing equipment used by manufacturers to test their cameras may give off some light, which can throw off the results. Another major challenge for low-light cameras is noise in the image captured in low-lux settings. All colors have something known as their own color temperature, which can produce noise. Noise from the color is the biggest issue for low-light cameras due to inability to perform absolute color reproduction, hence it is also important to have the digital noise reduction function to further improve image quality.

WDR ADDRESSES EXTREME LIGHTING

The wide dynamic range (WDR) feature is now standard for almost all security cameras as it helps to address lost details from extreme lighting contrasts present in complex indoor or outdoor environments. Users must note that dynamic range is defined as the ratio of the brightest and darkest regions in an image, and not as an absolute value.

The most common form of WDR uses the multi-exposure method, which consists of capturing two frames at one short and one long exposure speed. The first exposure captures details in bright areas in the scene while the latter captures details in dark areas in the scene. The two images are then combined for an image with visible details in both the bright and dark areas. Most common WDR falls somewhere between 50 to 70dB, while advanced WDR ranges from 100 to 130dB. Currently, the most advanced WDR, the third generation for that matter, is also called “True WDR," capturing four frames for comparison to achieve optimal results, as opposed to the two frames taken with past WDR technology. The increased number of frames taken for comparison helps to address issues such as lighting differences between the foreground and background in the image.

Though many manufacturers boast cameras with 130dB, in reality, it is already considered quite impressive if these cameras can reach around 120dB. Users must evaluate where they plan on installing their cameras and if WDR features are actually needed. Applications where WDR are typically needed include parking lots, big plazas, building entrances, ports, tollbooths, and other areas that are frequently affected by changing light conditions.

HIGHER FRAME RATES ENABLE FASTER-THAN-REAL-TIME VIEWING

Frame rate refers to the number of frames a camera can capture in a second. Human vision, in terms of real-time viewing, will equate to about 30 frames per second (fps). With current technologies, standards for frame rate have come to 720p HD (1,080 x 780 pixels) and 1,080p full HD (1,920 x 1,080 pixels) cameras capturing images at 30fps, providing users with real-time view in high definition. As technology continues to improve, the norm is now moving in the direction of 60fps at HD and megapixel resolution — faster than human vision. Increased frame rates with higher resolution will be extremely beneficial in applications such as toll roads and highways, where cars travel at a high speed, so law enforcement officials are still able to snap a clear picture of the license plates of any law-violating vehicles.

Not long ago, towards the end of 2013, LILIN came out with a 2-megapixel ip camera that is able to capture images at an astounding rate of 120fps. At present, the most common spec is 30fps for a 2-megapixel camera, even though there are megapixel cameras ranging from 1.3 to 29 megapixels. However, the higher the resolution, the slower the frame rate will be as the image sizes are larger and it becomes more difficult to compress with existing H.264 standards. Once H.265 transforms from myth to reality and starts being implemented on a wide scale, this issue will surely be resolved. A major deciding factor for frame rate is highly dependent on the CPU, which varies decisively even though they have the same core but it can offer multiple different features that lead to different performances. Most manufacturers prefer Arm processors, which are put in most SoCs of surveillance ip cameras.

SUFFICIENT BANDWIDTH FOR SMOOTH TRANSMISSION

Having sufficient bandwidth is extremely important for smooth streaming of videos; however, there are several factors that must be taken into account when calculating appropriate bandwidth needed for implementing an IP-based video surveillance solution system. These factors include recording resolution, number of cameras, compression type, fps, duration of recording, and if recording will be event-based or continuous. Also, compression type can be separated to MJPEG, MPEG-4, or H.264. The amount of bandwidth a video recording takes up is also dependent on the amount of motion that takes place in that environment. For instance, cameras can be configured to constant bit rate (CBR) or variable bit rate (VBR) mode. In busy settings where the scenery is constantly changing, such as bus terminals, train stations, airports, and so on, the cameras will be set to CBR. When recorded and played in CBR mode, the video footage is very smooth, but it can only be used for applications where there is enough bandwidth to support it or the image quality will be compromised. When recorded and played in VBR mode, bit rate only increases when abnormal activities or higher-than-usual activities occur. When files go over the bandwidth allowed, it will cause footage to lag. VBR is used in applications where changes to the scenery are rare and few, such as perimeter detection.

Users must note that neither of these settings is better than the other as long as they are used in suitable applications. The normal bit rate needed for HD 720p resolution is around 4Mbps and around 8Mbps for full HD 1,080p resolution. However, large video files often take up a large amount of bandwidth, and when bandwidth runs out, that is when problems arise. Common problems include video artefacts, dropped frames, and reduced resolution when transmitting data, as well as temporary loss of connection and freezing. This is another reason why cameras with ultra-high resolution are usually used primarily for identification and not for recording. To address this issue, users must learn how to properly manage their bandwidth so that it is carrying valuable data as opposed to redundant information.

STREAMING VIDEO: ONE WAY OR ANOTHER

Current video surveillance cameras are now able to provide up to three or four streams simultaneously. Usually, if a camera offers three streams, they are usually separated into one live, one recorded, and one remote stream. Advanced cameras are able to offer up to five streams, two live, two recorded, and one remote. The main stream is usually streamed and recorded in full HD with H.264 compression in real time at 30fps. The second stream will be recorded at a slightly lower rate, around 15fps. Depending on what the CPU can offer, the resolution of remaining streams will be lowered to CIF or D1.

At its best, a camera with three streams will be able to stream and record in real time at full HD resolution for all video streams. For anything higher than three streams, the stream rate and resolution will usually decrease for each additional stream, unless the CPU is powerful enough to handle more. As remote streams are able to be accessed and viewed through mobile devices, such as smartphones and tablets, they are oftentimes affected by services provided by telecom companies. When the files being viewed remotely are too large, it will be likely to lag.

Now, network cameras are already beginning to take over the video surveillance market. In a few years, the sales of network cameras are sure to surpass those of analog cameras because this technology is proving itself to be mature and reliable, while becoming increasingly easier to install and operate. Network cameras are at a point where they are able to record in high resolution in real time in a low-light environment, while addressing extreme lighting conditions, and stream these feeds in a stable and reliable manner. These network cameras are almost able to do what the human eye can do. The only thing manufacturers can do now is aim higher.