Challenge:

Enable real-time surveillance using mobile digital video recorder (MDVR) technology, even over the limited bandwidth available from 3G wireless connectivity.

Solution:

InfoDif Ltd. developed a highly efficient image-processing framework called Capra* and an MDVR called Falco*, based on that framework, to enable real-time surveillance. This technology enables real-time streaming and H.264 compression, even when using 3G wireless data transmission and small-footprint devices based on the Intel® Atom™ processor.

"With the help of Intel® developer tools, we first developed Capra* as a very high-performance image-processing framework that takes excellent advantage of multi-core processor architectures. That provided the foundation for the success we have achieved with Falco*, our mobile DVR solution."
- Mr. Kerem Caliskan, General Manager, InfoDif Ltd.

Customer Benefit:

With Falco, customers can provide very advanced mobile security solutions that deliver the high-performance, real-time surveillance necessary for functionality such as controlling a camera to view a specific visual area of interest. Those applications can easily transfer the recordings as needed and interoperate among the small portable devices and the more capable systems, based on Intel® Core™ i5 and Core™ i7 processors, and the Intel® Xeon® processor, for which the solution is also optimized.

Learn more: www.intel.com/partner

 

 

Falco*, a mobile digital video recorder (MDVR) from InfoDif Ltd. in Turkey, enables real-time video surveillance through advanced signal processing techniques it developed on Intel® architecture-based hardware.

 
 

A Breakthrough in Real-Time Video Surveillance

Security applications used in implementations as diverse as oilfields, airports, and convenience stores use video surveillance to monitor facilities and provide supporting security functionality. Operation requires robust, real-time video processing and transmission. Without that capability, monitoring cannot deliver many types of desirable functionality and protection:

• Support for camera movement. Users must have real-time feedback to control cameras from a central location, such as a security operations center, to avoid the possibility of losing the object of interest from view.
• Tracking transitory events. If a person, car, or other subject appears only briefly while passing by the camera, sufficient frames per second and resolution must be available to identify the event.
• Reporting of time-sensitive events. When personnel or valuable property are being protected, security events must be reported immediately to minimize the risk of injury or loss.

InfoDif is an engineering-driven company made up of computer scientists with expertise in video-signal processing and other related fields. With the increasing penetration and utilization of 3G wireless data infrastructures, the company saw the need for video surveillance applications that can support real-time video streaming and compression over 3G networks, and in Wi-Fi* or stand-alone operation.

To make the most effective use of its domain expertise, the company decided that providing a framework for such applications (as opposed to providing the applications themselves) was the soundest approach. To that end, InfoDif began work in 2008 on the Capra high-performance video-processing framework.

As one aspect of its strategy was to differentiate its product with very high performance, the company decided early on to build a very advanced threading model into its design. The efforts in that area have paid off with a highly efficient, robust signal-processing pipeline that scales on increasingly parallel processor architectures as they emerge.

To bring the functionality of Capra to providers of surveillance applications, InfoDif created Falco, a mobile digital video recorder (MDVR) solution. Falco is highly optimized for Intel® architecture, including the Intel Atom processor, making it well suited to the memory- and footprint-constrained devices that are in wide use in mobile security installations.

Extensibility as a First-Order Design Requirement
Because Falco is a platform on which security solution providers build their offerings, InfoDif values its adaptability to the needs of diverse solution areas. For example, some end-customers might require the ability to send very high-resolution video, while others require strong encryption of those feeds or very low use of bandwidth. Many solutions need a combination of these types of requirements, with the ability to tune the balance between them.

InfoDif built Falco with that extensibility in mind, and as a result solutions built on the platform have the capacity to customize requirements. For example, security solutions can utilize machine vision functionality developed by InfoDif that, when paired with optical character recognition solution components, could potentially be able to "read" license plates in search of a specific vehicle or to identify vehicles that enter an area under surveillance. These solutions might also use facial recognition to search crowds for individuals of interest, including those on watch lists, fugitives, or even missing children.

Other areas where extensibility plays an important role in security solutions include encrypting video streams, as mentioned above, which could be vital in protecting intellectual property. Those solutions might also need to be able to readily adopt alternate encryption, compression, input, or display modalities to interoperate with other solutions or business needs. InfoDif has placed a premium on that sort of extensibility (and the ability to provide it rapidly) throughout the development of Capra
and Falco.

Platform Extensibility to Protect Broad Applicability
In addition to providing framework extensibility, InfoDif recognizes the need to easily adapt to various Intel® hardware platforms. The unparalleled flexibility of the x86 architecture is crucially important in this area, providing an ecosystem of drivers and solution components that support very open-ended innovation:

• Backward compatibility. InfoDif’s use of Intel® Software Development Products such as compilers and libraries support the automatic tailoring of the solution’s behavior to the execution platform at runtime.
• Forward compatibility. As an Intel® Channel Partner, InfoDif has the opportunity early in the development cycle to test and optimize its solutions for emerging platforms.
• Readiness for increasing core counts. InfoDif’s ongoing relationship with Intel includes roadmap readiness as core counts on future processors continue to increase.

The diversity of test platforms that this approach embraces is very valuable to providers building security solutions on top of Falco. End-customer solutions can run on everything from lightweight embedded platforms based on the Intel Atom processor to workstations or servers based on the Intel Xeon processor. Regardless of the target system, InfoDif has prepared a technology foundation that makes excellent use of the hardware.

Maximizing Pipeline Throughput with Intel® Software Development Products

InfoDif engineers worked steadily to create a signal-processing pipeline robust enough to handle real-time video using mobile platforms. Throughout the development of Capra and Falco, they constantly strived to take full advantage of every aspect of the hardware’s capabilities to drive up performance. One key approach was to use Intel Software Development Products, which enabled them to readily benefit from the work already done by Intel performance engineers.

A Faster Track to Optimization with Intel® Performance Libraries
As the InfoDif engineering team surveyed the approaches and tools available to them to build robust parallel performance into its products, Intel Software Development Products represented a clear opportunity. Providing ready-made functions built to be dropped into software solutions, the libraries are comprehensive, highly optimized, and multi-core ready:

• Intel® Threading Building Blocks abstracted parallelism away from program logic, enabling InfoDif engineers to focus on value-added domain expertise instead of multi-threading.
• Intel® Integrated Performance Primitives delivered pre-optimized algorithms for common functionality in areas such as signal-processing, compression, and encryption, simplifying the delivery of robust parallel performance.
• Intel® Math Kernel Library gave the InfoDif engineers access to vector and matrix math operations that were pre-tuned to deliver excellent performance on Intel platforms.

Simplified Tuning with Analysis Tools from Intel
With the development of its core pipeline and supporting functionality in place, the InfoDif team turned its attention to optimizing the performance of the broader solution. That effort included the identification of bottlenecks and other opportunities for improving performance, iteratively at various stages of the application development life cycle. This simple approach, enabled by analysis tools from Intel, required little or no instrumentation or recompilation:

• Intel® VTune™ Performance Analyzer helped the team rapidly identify hotspots where the framework was using intensive processor resources, as areas of interest for possible optimization and also gave ideas for pursuing further tuning, down to the code-line level.
• Intel® Thread Checker helped the team refine its threading model to take optimal advantage of parallel processing resources, identifying issues such as race conditions and excessive thread synchronization that could have reduced performance or even introduced errors during operation of the framework.

Simple Generation of High-Performance Binaries with the Intel® Compiler
An important contributor to InfoDif’s generation of high-performance, flexible code, the Intel® C++ Compiler is built by Intel performance engineers who have an intimate knowledge of the hardware. That unique expertise contributed to the dramatic performance leaps made by the Capra and Falco teams throughout development.

A robust set of compiler switches were also available to the developers, as well as a world-class body of documentation, white papers, user forums, and other support. The team used these resources to excellent advantage, as demonstrated by the final result of their efforts—the foundation for the industry to build the next generation of real-time mobile video surveillance solutions.

Building a Secure Future

As the InfoDif engineering team continues to innovate, they have developed methodologies that enable them to very quickly customize their framework to the needs of individual customers. In a rapidfire industry where providers demand immediate access to open-ended innovation, InfoDif continues to meet the challenges.

Learn more about InfoDif Ltd: www.infodif.com

Visit the Intel® Software Network Parallel Programming Community: software.intel.com/en-us/parallel

With a flexible framework ready to expand on demand to incorporate new functionality and additional platforms, Falco is ready to support the imaginations and needs of security solution providers across the spectrum. As new usage models, devices, and security needs arise, InfoDif is ready to help protect facilities, people, and operations of all kinds, all over the world.

About the Intel® Software Partner Program
The Intel® Software Partner Program provides a framework for collaborative solution development around Intel® architecture. From business planning and product development to marketing and sales, the program helps to drive increased business success and market opportunities. Learn more at http://software.intel.com/partner.

 

Intel® compilers, associated libraries and associated development tools may include or utilize options that optimize for instruction sets that are available in both Intel® and non-Intel microprocessors (for example SIMD instruction sets), but do not optimize equally for non-Intel microprocessors. In addition, certain compiler options for Intel compilers, including some that are not specific to Intel micro-architecture, are reserved for Intel microprocessors. For a detailed description of Intel compiler options, including the instruction sets and specific microprocessors they implicate, please refer to the “Intel® Compiler User and Reference Guides” under “Compiler Options." Many library routines that are part of Intel® compiler products are more highly optimized for Intel microprocessors than for other microprocessors. While the compilers and libraries in Intel® compiler products offer optimizations for both Intel and Intel-compatible microprocessors, depending on the options you select, your code and other factors, you likely will get extra performance on Intel microprocessors.

Intel® compilers, associated libraries and associated development tools may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include Intel® Streaming SIMD Extensions 2 (Intel® SSE2), Intel® Streaming SIMD Extensions 3 (Intel® SSE3), and Supplemental Streaming SIMD Extensions 3 (Intel® SSSE3) instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors.

While Intel believes our compilers and libraries are excellent choices to assist in obtaining the best performance on Intel® and non-Intel microprocessors, Intel recommends that you evaluate other compilers and libraries to determine which best meet your requirements. We hope to win your business by striving to offer the best performance of any compiler or library; please let us know if you find we do not.

Notice revision #20101101