Sigma Provides Image Pipeline Processing Using Blackfin® For Digital SLR
Capturing original optical data and weaving it into shapes with beautiful colors is what every digital camera does. So says Sigma, a Japanese-based company that manufactures digital cameras, lenses, and other photography products. For Sigma, creating an exceptional digital SLR (single lens reflex) camera that portrays natural beauty, and accurately captures light without compromises is an ongoing challenge. Sigma's drive for image excellence had previously led the company to adopt an advanced, direct-image sensor technology from Foveon in its SLR camera family.
For its next camera, Sigma wanted to push its image quality a step forward with a new 14-megapixel Foveon X3® sensor enhanced by software-based image pipeline signal processing. Sigma selected a Blackfin® ADSP-BF561 processor from Analog Devices, Inc. (ADI) for the task because the processor provided both high performance and low power consumption. The other signal processors Sigma researched could not match Blackfin's attributes.
Powered by the unique direct-image capabilities of the Foveon X3 sensor and a Blackfin ADSP-BF561 processor (for real-time image-pipeline number crunching), Sigma's new SD14 Digital SLR camera can reproduce high-definition images that are rich in gradation and provide impressive 3-dimensional detail. Most digital cameras use interpolated, mosaic-based image sensors that only capture one-third of the color data, introducing color artifacts and resulting in a loss of image detail. By contrast, the vertically stacked Foveon sensor directly captures red, green and blue data for each and every pixel, without loss or distortion.
As with all digital cameras, users of the Sigma SD14 can choose to save images in compressed JPEG mode. But when photographers want complete control of their artistic expression, they can record RAW, unprocessed image files. RAW files retain the most complete set of data, produce the highest image quality, and result in the most accurate rendering of subject color and texture.
Sigma's SD14 Digital SLR Camera
Math-Intensive Image Processing
RAW image data collected from the Sigma SD14 camera's sensor streams into the image pipeline where a Blackfin ADSP-BF561 processor executes complex image-processing algorithms. This typically involves image optimization steps like dark current subtraction, flare compensation, shading and color compensation, demosaicing, white balancing, tonal and color correction, sharpening, and compression.
The entire image pipeline, from sensor to memory card, must execute very quickly in real-time to ready the camera for the next shot. This is no small task since image pipelines for high-quality SLR digital cameras might perform as many as 200 computations per pixel for each and every image. Without a high-speed processor on board, the shot-to-shot computational time required for a 14-megapixel camera would be unacceptable for most camera users. Fortunately, the Blackfin ADSP-BF561 processor, with dual 600 MHz symmetric processor cores, provides the high performance required to accelerate image pipeline processing for Sigma's new camera.
Because Blackfin is based on a programmable architecture, Sigma was also able to optimize its image processing algorithms for maximum impact on the SD14's image quality. The algorithms ensure accurate color, remove common types of lens distortion, balance the light levels inside the camera to match the incoming light, adjust the relative brightness, sharpen edge detail, enhance the visibility of a boundary between light and dark tones in an image, and more.
The main reason Sigma chose a Blackfin ADSP-BF561 processor was because the processor offered a good balance between high performance and low power consumption. A high-performance member of the Blackfin family, the ADSP-BF561 processor comprises two independent Blackfin 600 MHz cores. Because the processor employs a dual-core symmetric multiprocessor instead of a single-core processor, performance is increased without switching processor architectures. This kind of power is ideal for mathematically intensive multimedia applications such as image processing.
For development tools, Sigma used ADI's VisualDSP++ integrated development and debugging environment (IDDE) to optimize its image processing algorithms. VisualDSP++ enables efficient management of projects from start to finish from within a single interface. It allows developers to move quickly and easily between editing, building, and debugging activities. Key features of VisualDSP++ include native C/C++ compilers, advanced graphical plotting tools, statistical profiling, and the VisualDSP++ Kernel (VDK), which allows code to be implemented in a structured easy-to-scale manner.
Another key reason Sigma chose a Blackfin processor was for its low power attributes. When portable battery-powered applications, such as Sigma's SD14 SLR Digital camera, need to run for a long period of time, the low power states of Blackfin processors can be very effective. By using the on-chip power management features such as the programmable voltage regulator, PLL, and low power modes, developers can maximize battery life by using only as much processing power as required.
Blackfin ADSP-BF561 processors provide additional options for power management because the processor contains two identical cores. Dual-core Blackfin processors contain large amounts of on-chip memory 328K bytes along with data paths and direct memory access (DMA) controllers sized to handle shared processing loads. As such, traditional processing-intensive applications can be split equally to run on each of the two cores with no loss of efficiency. Applications can therefore run at half the frequency of a single-core system, and the processor core voltages can be dropped.
Without question, the Sigma SD14 Digital SLR camera, with its three-layer, full-color image sensor, is a groundbreaking product. And with the image processing muscle of Blackfin, this digital camera is about to set a whole new standard for image quality.
For more information, please visit Sigma.
Read more Customer Case Studies like this...