How to Succeed in the HD Network Camera Market

Sunell shares its expertise on network camera selection.

In the high definition (HD) era, the market has placed increased demands on surveillance products. While analog video users asked for clear images, today they request crisp detail as well as realistic color saturation. Low-light performance must not just be as low as possible, but wide dynamic range (WDR) should have outstanding results as well. As higher resolution and larger camera networks are the trend, products are evolving. Resolution is becoming ever larger, from 720P, 1080P up to the 5-megapixel range. Low illumination and WDR are also becoming requirements for the HD market.

The problem facing manufacturers today is not the lack of market demand, but rather the inability to meet customer demands. In the HD era, product development is no longer about buying a Sony camera module, putting in an enclosure and then shipping it. The technical threshold has been raised, above the abilities of traditional camera makers. As high resolution quality products are the norm, vendors who possess component know-how will have an edge in the market.

Image quality is the greatest differentiator for HD video. No matter how stylish the camera design is or how powerful the features are, good image quality cannot be lacking. Outstanding image quality begins with good hardware, with the image sensor being the deciding component. Even the best image processing cannot clean up images from an underperforming image sensor.

How to Select Image Sensors
Image sensors are either CCD or CMOS, which differ in their architectural design. For CCDs, each pixel is connected to an electrical “temporary storage,” which converts photons to electronic signals after exposure to light. These signals will be activated by TG's H and V drivers and then transferred to the ADC, which amplifies and converts the signals into digital ones. As CCDs use charge transfer method, these require to set up specific transmission channels, making them expensive and challenging to produce.

By contrast, each CMOS pixel is connected to an ADC, so photons are directly converted into digital signals. Each ADC amplifies the signal differently, making it hard to achieve a simultaneous conversion. Compared to the one-time amplification of CCD signals, CMOS images are noisier, particularly under low-light conditions. Each CMOS pixel includes an amplifier and A/D converter, the additonal equipment over a single pixel surface reduce the light-sensitive area size, so a CMOS sensor with the same format as a CCD one will have worse low-light performance.

CMOS sensors today show marked improvement in noise reduction, light sensitivity and low-light performance, coming close to CCD levels. CMOS production costs are also lower than CCDs, which is another advantage. Thus, CMOS specialists such as Aptina are gaining increased market recognition. Even Sony, the reigning champion of CCD production, is developing CMOS sensors. As video surveillance migrates from analog toward network HD, CMOS will become more dominant in the market.

Select sensor, the first factor is the light-sensitive area size of single pixel. While a larger image sensor will have larger pixels that are more sensitive to light, this benefits for low light perforamnce and noise suppression, but it will take up more space in the camera and use more power. This affects product design and heat dissipation. Large-format sensors also rarely have matching lenses, making it hard to find the right lenses for 2/3” to more than 1” sensors.

Noise reduction and WDR are being developed for more sensors, such as Pixim's DPS dynamic range function. Pixim partnered with Sunell for a super WDR color camera that delivers strong performance. Aptina has also developed a surveillance-specific WDR sensor integrated with a DSP that can process WDR value up to 120 decibels.

Finding the Right Solution
Options for high resolution imaging are limited. The first is using an image sensor integrated with an image sensor processor (ISP), which outputs YUV data and compresses into H.264. This option has limited customization options and image quality will be largely the same.

The other option requires stronger R&D ability. Some image sensors do not include an ISP and output RAW data. This data format requires a DSP or FPGA to act as the ISP. However, FPGAs cost more than ISPs. Sunell's programming expertise overcomes the limitations of the 2A algorithm for superior image quality.

Integrated Cameras
There is increased demand for integrated cameras, which feature PTZ and IR capabilities in one camera. These cameras place unique demands on image sensors, with more product offerings from Sony, Hitachi, LG, CNB, Sanyo, Hitron and other manufacturers.Integrated cameras are suited for many verticals, as well as for nonsecurity purposes such as teleconferencing. Sunell is one of the first Chinese manufacturers to launch an integrated camera with auto focus for greater clarity and detail. A manufacturer's core competence in component R&D will determine its competitiveness.

Share to:
Comments ( 0 )
Most Viewed Articles