Popping the Hood on 180- & 360-Degree Cameras
While quality and performance should be the main considerations when selling and installing video surveillance systems, the importance of cost to customers cannot be dismissed. This fact is one reason that 180- and 360-degree cameras have gained traction lately. Offering the potential of lower equipment, installation, maintenance and other costs with reduced camera counts, these cameras have become an attractive option for a variety of customers — whether they are budget-conscious or not. The quality and performance that 180- and 360-degree cameras offer doesn’t hurt their popularity either.
To provide dealers and integrators with the ability to make the best choices for their customers, SDM is taking a look under the hood of these cameras to see how they work, and to examine the pros and cons of the various technologies they employ.
Typically, 180- and 360-degree cameras are based on two hardware platforms: single-lens/single-sensor and multi-lens/multi-sensor. In a single-sensor camera, the imaging sensor works in combination with either a fisheye or panomorph lens to capture images, which must then be de-warped to produce either a 180- or 360-degree view.
A multi-sensor camera incorporates up to four imaging sensors, an equal number of lenses and image-processing capability within the same housing. For example, an eight-megapixel camera would consist of four separate two-megapixel modules. The images gathered by the sensors are then stitched together to generate a full field of view, but that isn’t the only approach, says Patrik Pettersson, product analyst, North America, Axis Communications, located in Chelmsford, Mass.
“A camera capable of 180 and 360 degrees may not always be used to cover those exact fields of view. Rather, they may leverage the multi-sensor capability in order to individually aim each sensor and lens combination to see specific areas of interest that are not in perfect 180- or 360-degree field of view,” he explains. “Or they can be configured for an outside corner of a building with the purpose of covering the 270 degrees of that corner, for example.”
Sales & Installation Tips
Given the specialized nature of 180- and 360-degree cameras, there are a number of factors to consider when selling or installing them. Below are some tips from industry professionals with good knowledge of these cameras to help integrators choose the right cameras — and maximize their performance — for their customers.
Heidi Barnum, MOBOTIX: “Fisheye cameras should include onboard de-warping technology so that the end user gets the most usability out of eacha and every image. Also, when selling hemispheric cameras, be sure to emphasize the fact that the customer’s camera counts will likely be lower. This not only saves the customer money now; it also keeps maintenance costs low over the life of the system.”
Allen Chan, Sony:“Be realistic. Just because you can see everywhere doesn’t mean you can recognize smaller details. On the flip side, often when you have access to smaller details, the range of field is not as extensive. There are limitations to effective resolution based on distance, vantage point and type of lens and you need to recognize that. Each type of product has a different use and area it is most effective in, so know the specific features and build a comprehensive security plan based upon that.”
Jeff Corrall, March Networks: “Consider the height at which a camera can be installed. Some work better when they’re higher up, and others are designed for eight- to 12-foot ceilings. The proper lens-sensor combination ensures very good focus on images at the edges.”
Mark Espenscheid, Digital Watchdog: Be aware of the loss of resolution in a de-warped image. Use panoramic cameras in your design for wide area surveillance, and then add single-sensor cameras to cover choke points for face and plate recognition.
Ahmed Elsayed, Hikvision USA: “We recommend that dealers and integrators really do their homework — learn about the limitations and features of the cameras — in order to set realistic expectations. While these cameras are loaded with outstanding features, they’re not a ‘one size fits all’ solution.”
Steve Gorski, Scallop Imaging: “Many 180- and 360-degree cameras can be difficult and complex to set up. Installers should look for manufacturers that offer products that can be easily installed and configured and are open-architecture.”
Patrik Pettersson, Axis Communications: “Use tools such as Google Maps and Sketch Up to give the customer a real preview of the type of coverage the camera will provide based on the standard. One-eighty and 360 cameras are great complements to existing installations and future systems when combined with pan/tilt/zoom and fixed cameras, but make sure end user expectations are met by utilizing the tools mentioned above.”
Tom Cook, Samsung Techwin America: “Don’t oversell the capabilities of the camera. Demo the camera to verify what the field-of-view and resolution will be. For a wall-mounted 180-degree camera, make sure to use a tilted bracket to ensure that you are not looking at the sky. For ceiling-mounted cameras, lower the camera as much as possible.”
Of the two sensor models, the fisheye/panomorph approach tends to be more widely deployed, says Tom Cook, vice president of sales for Ridgefield Park, N.J.-based Samsung Techwin America.
“It’s relatively easy to place multiple imagers into a housing to create a panoramic camera. Unfortunately, you then essentially have multiple cameras to adjust, focus and configure. Some recording platforms would see these as separate cameras, adding to the hassle and costs of camera licenses,” he says. “Now that higher-megapixel cameras can be efficiently transmitted and recorded, single-sensor panoramic cameras are becoming very popular.”
However, the image quality produced by single-sensor/single-lens cameras can be hampered by a number of issues, including limited depth-of-field, pixel density at the edges of the lens, and poor low-light sensitivity with higher-megapixel resolution, Pettersson says.
“They are not an ideal choice for a camera reduction strategy and are improperly used as an identification product when they should really be used as a complementary overview camera in most applications,” he says. “In some cases, they require good integration with video management software for de-warping purposes.”
One potential misunderstanding about 180- and 360-degree cameras is that they aren’t always separate technologies. A single-lens, single-sensor 360-degree camera can be easily transformed into a 180-degree camera by wall mounting rather than ceiling mounting it, Pettersson says.
“Top down, the camera is operating in a 360-degree capacity; when mounted on a wall — with a bit of digital cropping from the top and bottom of the image — you can transform it into a 180-degree camera,” he describes. “The single-lens 180/360 camera also covers the entire field of view, even dead-center underneath the camera, whereas a multi-sensory, multi-lens camera has a void underneath because its four lenses point out from the center core of the product, which can be adjusted by pointing the lenses down. But this setup causes loss of critical depth-of-field. The 180/360 multi-lens cameras are better at covering larger areas due to their native capability of a longer depth-of-field.”
A main difference between the two sensor approaches is that multi-sensor cameras are used more for longer views rather than capturing images directly below the camera.
“Multi-sensor cameras are generally used when you want to cover a large area in a significant amount of detail. A 360-degree single sensor is used for a smaller area to reduce camera counts or cover the entire room,” says Jeff Corrall, product manager, Edge Devices for Ottawa-based March Networks. “Too often, these cameras are lumped into the same category. You may require a 360 but don’t know whether you want a multi-sensor or single-sensor because the differences are so vast.”
In addition to lower costs, single-sensor cameras also tend to be easier to install and have a lower risk of component failure. In terms of practical applications, single-sensor cameras have limited depth-of-field, making them well-suited for providing situational awareness at shorter distances, such as mounting above a cashier or teller to monitor transactions.
“There are pros and cons to each approach. If you are using a single camera, you are using only one stream versus larger multiples of streams, which present bandwidth and storage issues to consider,” says Allen Chan, senior product manager, security, Sony Electronics, Park Ridge, N.J. “Multi-sensor is utilized more for looking farther out, as the vertical field of view is limited, while fisheye/panomorph allows you to see all around and is used more for situational awareness, including surveillance of a known entity — for example, monitoring someone on suicide watch.”
Past performance in particular applications can be a good indicator of how 180- and 360-degree cameras can be used. Below are four examples of installations where these cameras were used to solve a security or other problem.
In Bergen County, N.J., MOBOTIX hemispheric cameras are used to provide video coverage of interior and exterior portions of the county jail, as well as the holding cell located in the courthouse. The estimated number of traditional cameras required for the system was significantly reduced by using wall-mounted hemispheric cameras, which provide the same amount and quality of coverage as multiple traditional cameras.
Albuquerque’s Hope Christian Academy, which encompasses a high school and middle school, has a 180/360 in its gymnasium. The panoramic view sees the entire gym, and each of the digital PTZs is assigned to different school administrators. The digital PTZ system lets users use the camera like a conventional dome camera without any moving parts, improving reliability.?Because there are different groups of students using the gym throughout the day, the principal of the high school has access to his own PTZ, as do the vice principal of the high school and the principal of the middle school. Each user can manipulate his or her PTZ remotely and autonomously without stepping on each other’s toes, while also viewing the 180 to see the big picture.
A customer of ENE Systems (a Scallop Imaging channel partner) had an existing surveillance system that employed a number of PTZs to secure its parking lot. Unfortunately, these cameras routinely seemed to be positioned somewhere other than where an incident occurred, and the customer was rarely, if ever, able to capture any of the incidents. By replacing these with a Scallop Imaging 180-degree camera, the customer is now able to view the entire parking area, which ensures that all incidents will be captured.
One of March Networks’ hotel customers wanted to provide surveillance of its business center, which was located in a very small room. As a result, there was nowhere to install a single traditional camera to cover the entire room because of blind spots, meaning multiple cameras would be needed. By installing a 360-degree camera, the hotel solved the problem of providing video coverage of the entire room while saving money by reducing its camera count.
Single-sensor cameras also can suffer from inherent pixel loss based mainly on a problem of geometry, says Steve Gorski, chief sales officer for Boston-based Scallop Imaging, adding that de-warping can further compound that issue.
“The biggest disadvantage with single fisheye lenses is the significant waste and loss of pixels, up to as much as 75 percent. A camera using a fisheye lens typically projects a circular image on a rectangular image sensor. That alone results in 50 percent of the pixels being lost,” he says. “In order to create an image the human brain can make sense of, fisheye images then have to be de-warped. This de-warping process results in another 25 percent of the pixels being lost.”
Multi-sensor cameras share many of the same pros and cons with their single-sensor relatives, with a few key differences, Pettersson explains.
“These [differences] include benefits such as higher-quality lenses that provide greater image quality at longer distances and a larger platform that gives manufacturers the ability to add more capabilities onto the camera. But of course, these benefits come with a higher price tag when compared to its counterpart,” he says.
Another benefit of single-sensor cameras that may be overlooked is based on form factor. Because they don’t look like traditional security cameras, they can be placed in locations that were previously inaccessible, such as museums, hospitality, high-end residential, casinos and elevators, Cook says.
“For retail locations, you can cover many aisles with one camera, and a criminal casing the joint can’t simply look up to determine where the camera is pointed,” he says. “Multi-imager cameras can be useful for locations with fixed views, such as four-way hallway intersection to replace multiple cameras.”
The images generated by hemispheric views are convex, particularly at their borders. So once images are captured, they must be processed to produce a view that humans can comprehend more easily. The de-warping process employed by single-sensor cameras applies software to the raw image to correct for curvature to allow viewing of the scene from the usual perspective.
This processing can be performed either on the camera itself prior to transmission or on the client side using software once video has been transmitted to the head end. For the most part, figuring out the best approach depends on who you ask, which doesn’t make sorting out the right choice very easy.
The benefits of camera-based de-warping start with the compression that is applied to the video prior to transmission, which some experts say is much better than with client-based de-warping.
“It is very important that the de-warping take place on the raw image before compression. This ensures good image quality in all parts of the image, specifically at the edge,” says Heidi Barnum, head of internal operations, MOBOTIX Corp., located in New York.
Conversely, having access to the raw image allows for more functionality, Corrall says. “With client-based de-warping, you’re technically streaming and recording at full resolution, which allows digital PTZ in the full image,” he says. “Customers are forced to pre-select the desired field-of-view they want to see when de-warping is performed in the camera. They can select up to four views to send to the VMS, but once you de-warp at the camera, there’s no going back to other views — and there’s no way to perform digital PTZ.”
Because there is no clear-cut “right way” to perform de-warping, many manufacturers are starting to recognize the value of both approaches and, as such, are diversifying their product ranges. For example, while Hikvision’s cameras available in North America currently employ software-based de-warping, the company doesn’t dismiss image processing within the camera.
“We also promote on-board de-warping, since the video is just like any other view and is recorded as such without any special integration required,” says Bob Germain, senior product manager at City of Industry, Calif.-based Hikvision USA. “This provides great flexibility when being used with third-party video management software. It is also useful on NVRs in terms of providing direct access to the de-warped stream at the NVR itself without the software.”
Starting with the potential to reduce camera counts, the benefits offered by 180- and 360-degree cameras are numerous. When evaluating the various models and manufacturers, the technologies cameras employ to capture and process images are the main factors to consider when evaluating 180- and 360-degree cameras. Each has its benefits and drawbacks, so it’s crucial that integrators be as informed as possible on the differences before making the right choice for a customer application.