What’s STARVIS/STARVIS 2?
One of the essential features for security cameras is the imaging ability that delivers clear images in dark locations. Sony’s image sensors with STARVIS/STARVIS 2 technology have the sensitivity well above that of the human eye. They deliver excellent performance in the low- illumination monitoringas they not only capture the targets in dark places, but even visualize their textures clearly.
Both STARVIS and STARVIS 2 are back-illuminated pixel technology specifically developed for CMOS image sensors for security camera applications. They feature a minimum sensitivity of 2,000 mV/μm2 (color product, when imaging with a light source of 706 cd/m2, F5.6 in 1s accumulation equivalent) and deliver high image quality in the visible light as well as near infrared light regions. STARVIS 2 further offers the high dynamic range (AD 12bit) of more than 8 dB in a single exposure, wider than the STARVIS pixel of the same size.
What’s STARVIS?
Achieving high sensitivity
Unlike front-illuminated image sensors, the back-illuminated image sensors with STARVIS/STARVIS 2 can collect more light onto the photodiodes unobstructed by wiring or circuits, achieving high sensitivity. (See Fig. 1)
The pixels specially developed for security camera applications can capture clear images in the low light environment, with the details that human eyes and conventional security cameras could not ascertain, such as objects’ colors and shapes, textures, even printed figures on a scale. (See Photo 1)
[Fig. 1] Advantages of the back-illuminated structure
[Photo 1] Performance comparison in low illuminance environment (illuminance 0.08 lx)
Image sensors with STARVIS/STARVIS 2 technology can capture images in the low light environment with details such as colors, shapes, and textures of the objects, which conventional front-illuminated image sensors are unable to.
- Conventional
Front-illuminated image sensor
- Sensor with STARVIS technology
Back-illuminated image sensor
Type 1/2.8, F=1.4, Exp. 1/60s, Frame rate 30 fps, Gain 48dB, With IRCF
Imaging performance in the near infrared (NIR) region
While traditional cameras require the visible light environment for shooting, security cameras may require imaging in the near infrared (NIR) light spectrum. All the existing models of Sony’s image sensors for security come with the high-quality NIR imaging feature, enabling shooting by NIR radiation in complete darkness.
[Photo 2] shows a comparison between images taken under the 0 lx condition: one by visible light type image sensor, and the other by NIR-compatible image sensor.
Sony continues to improve the NIR feature of STARVIS. The new model, IMX585 has been released recently, offering better imaging performance with even lower noise than its predecessors.
* Find our products with improved NIR imaging feature in the product list (Select NIR Enhanced).
*[Photo 3] shows a comparison of NIR imaging features between the previous and improved versions.
[Photo 2] Comparison against visible light type image sensor
Under the lighting condition of 0 lx illuminance with 850 nm NIR light, the NIR-compatible image sensor with STARVIS 2 IMX585 can capture not only the shapes of the objects, but also shades and details such as the letters printed on the scale, while the image sensor for visible light cannot produce an image.
- Image Sensor for visible light
(Not compatible with NIR sensitivity)
- Image sensor with high NIR imaging performance
IMX585
Condition : F=1.6, Exp.=33.3msec, Gain=0dB, 850 nm
[Photo 3] Comparison between the conventional and new versions of NIR imaging feature
- NIR compatible sensor (conventional version)
IMX415
- NIR compatible sensor (improved version)
IMX515
Condition : F=1.6, Exp.=33.3msec, Gain=0dB, 850 nm
High Dynamic Range (HDR)
Fixed-position security cameras need to adjust themselves to the lighting conditions that change over time and/or by the changes in the environment. If the scene has both too bright and too dark regions at the same time, the bright region may be overexposed while the dark region may be blacked out. To ensure objects in either regions be captured clearly, there is a method to adjust the exposure time or gain to avoid blowout or blackout, called high dynamic range (HDR) feature. (See [Photo 4])
[Photo 4] Blowout reduction using HDR feature
- HDR off
- HDR on
- The background is too bright, causing blowout. The target is not discernible.
- The blowout effect is reduced. The target color and shape can be discerned.
Currently, there are two types of HDR feature for STARVIS/STARVIS 2: DOL HDR (Digital Overlap High Dynamic Range) and Clear HDR. STARVIS is compatible with DOL HDR only while STARVIS 2 is compatible with both DOL HDR and Clear HDR.
DOL HDR
When the DOL HDR feature is turned on, the image sensor captures two images in quick succession, one with a long exposure according to the dark region, and the other with a short exposure adjusted to the bright region. The output data are synthesized in the post processing (Note 1). By synthesizing two images of different exposures, adjusted to the dark and bright conditions, the sensor can facilitate a wider dynamic range than a single exposure image. This method can be employed effectively for capturing still targets. Note, however, that the slight time lapse between the two shots can cause some artifacts, such as chromatic aberration, if the target is in fast motion (see the left image of [Photo 5]).
Clear HDR
When the Clear HDR feature is turned on, the image sensor captures two images simultaneously, one with a high gain set to the dark region, and the other with a low gain adjusted to the bright region. The output data are synthesized in the post processing (Note 1). By synthesizing two images of different gains, configured for the dark and bright conditions, the sensor can facilitate a wider dynamic range than a single exposure image. As the two images are shot at the same time by this method, moving targets do not cause artifacts in the resulting image (without chromatic aberration, etc., see the right image of [Photo 5]), suitable for AI processing, etc.
(Note 1) To obtain the final image, the output data must be processed by the camera. IMX585 can synthesize data internally when using Clear HDR.
[Photo 5] Comparison of images with a fast-moving target between DOL HDR and Clear HDR
The two images below illustrate a comparison between DOL HDR and Clear HDR, capturing the same metronome with a moving part. In both images, the details in the bright background and the letter “S” on the metronome placed in the dark area are clearly captured.
However, the illustration attached on the pendulum shows some chromatic aberration in the image taken with DOL HDR. The image taken with Clear HDR does not have chromatic aberration.
- DOL HDR
- Clear HDR
Provision of STARVIS/ STARVIS 2 logo data
Customers may use the “STARVIS/ STARVIS 2” logo with Sony’s approval in the following circumstances only. Before provision of the logo data, a logo usage agreement will be concluded with the customer. Please contact a sales office in your area for details.
Products: customers’ products incorporating Sony’s CMOS image sensors equipped with STARVIS/ STARVIS 2 technology.
Materials/Media: product packaging, catalogs, posters, informational websites and other promotional materials, as well as fixtures at exhibitions (panels, etc.) relating to the above mentioned products.
Source: https://www.sony-semicon.co.jp/e/products/IS/security/technology.html
