Semiconductor-based imaging sensors are coming on strong today, thanks to new advanced features, reduced power consumption, and lower costs. Design engineers are increasingly finding ways to apply the technology in a whole range of products, including cell phones, digital still cameras, PC cameras, cars, security and surveillance systems, bar code readers, medical devices, and even toys.
Image sensors are based on either CCD (charge-coupled device) or CMOS technology. High-performance imaging sensors (with resolutions of 5 megapixels or more) rely on CCD semiconductor technology, while low-cost imaging has historically been associated with CMOS technology.
However, the lines have begun to blur now that advancements in CMOS have extended its capability to up to 5 megapixels.
High and ultra-high performance applications continue to rely on CCD technology. Omron's F160-S2 double-speed camera, for example, can perform measurements as fast as 3 msec. Likewise, Keyence's CV-2600 Series camera provides ultra-high accuracy detection with a 2 million pixel CCD. Sub-pixel processing and image digitization offers repeatability down to ±0.05 pixels.
Advancements include increased integration at both the silicon and package level. For example, Phillips Semiconductor integrates a Video Graphics Array camera, a TFT display and driver, and its Nexperia mobile image processor into a single unit.
A typical imaging application requires three critical elements: the imaging sensor, an image processor, and an application processor. Some suppliers, such as STMicroelectronics make both the image sensor and the image processor—a chipset for mobile imaging. ST's STV0976 digital image processor supports its 1 megapixel Standard Mobile Imaging Architecture (SMIA) camera module at up to 15 frames per second (fps), or up to 30 fps with its Super Video Graphic Adapter (SVGA) unit.
Other suppliers provide dedicated image processing chips. For digital still cameras, TMS320DM270 processor from Texas Instruments integrates an 80 MHz ARM7 32-bit processor, a 90MHz TI C54x digital signal processor, and a 180 MHz single instruction, multiple data (SIMD) image processing engine for programmable image processing to support both CCD and CMOS image sensors. The engineer designing a vision system has several choices—what works best for a particular application will depend on the system performance and cost objectives.
The acompanying table compares some of the latest offerings in vision sensors and systems.
The first megapixel mobile-camera chipset compliant with the Standard Mobile Imaging Architecture (SMIA)
Digital still camera
3 megapixels
2,000 × 1,500 pixels active array
6.75 mm × 5.08 mm sensor area
Autobrite 120 dB wide dynamic range for adapting to varying light conditions, electronic rolling shutter, on-chip 12-bit column parallel analog to digital converter
140 mW; Standby mode, <0.5 mW
-10 to +50C
1× to 16× digital gain, 8 or 12-bit digital output
High level of integration in ASIC and sensor, and autobrite wide dynamic range and longevity low power design
Very small, thin VGA camera modules designed for camera-enabled compact cell phones. Unit incorporates a double lens and a single chip integrating a CMOS image sensor and a digital signal processor.
Cellular telephones, PDAs, and other portable equipment
330,000-pixel VGA
660 × 492 array of pixels with 1/6 of an inch optical format
6 mm (W) × 6 mm (D) × 4.5 mm (H)
Double lens and a single chip integrating a CMOS image sensor and a digital signal processor
40 mW at 30fps
-20 to +60C
Digital outputs: YUV = 4:2:2 or RGB Scanning: Progressive-scanning Frame rate: Maximum 30 frames per sec
Individual 3.75-microns pixels allows approximately 330,000 signal pixels on a single chip with a 1/6 of an inch optical format
$20 in 10K quantities
Web Resources
//Check out the links below for additional imaging resources//
Image sensors are based on either CCD (charge-coupled device) or CMOS technology. In a CCD sensor, the output is an analog signal. Every pixel's charge is transferred internally through a limited number of output nodes, converted to a voltage signal, and buffered. In a CMOS sensor, the output is digital with a charge-to-voltage conversion for each pixel. The sensor frequently includes additional digital circuitry.
JUNE 26TH WEBCAST: Collaborative Requirements Engineering
Speed your innovation. Capture the "voice of the customer" and translate customer requests into user requirements that define new products. Find out why the new ENOVIA Requirements Management solution enables organizations to improve their overall global requirements management process. Read More
Mechatronics in action
Successful synergistic integration of controls, electronics, computers and mechanical systems is key to the 21st century design process. Unlock the secrets at the Mechatronics Zone!
Webcast: Sensor Know-How Now
Join our moderator Randy Frank and John Keating from Cognex and explore Solving Industrial Inspection Problems. Read More
Engineering Concept Conduit
Engineering Concept Conduit looks at new products and the components that make them exceptional. Each month we’ll look at a new electronic product and see what makes it tick from an engineering point of view. We’ll explore the design and engineering challenges for the product and examine the components that solved those challenges.
Light Matters: Systems Level Approach to HBLED illumination applications
Its good practice to apply a systems-level approach to high-brightness LED (HBLED) illumination applications. Minimally, the system includes the optical, thermal and electrical characteristics of the of the HBLED, the lens (if any) which is built-in to its package, secondary optics such as external plastic lenses/reflectors to direct the light as your application requires and power driver electronics. Read More
Design News RSS News Feeds
Webcast: Sensor Know-How Now
