To meet the market's demand for small area scan cameras with both high resolution and high image quality, JAI has selected the Pregius S series – the latest generation of CMOS image sensors from Sony.
This 4th generation of Sony CMOS sensors represents the state-of-the-art in pixel technology for machine vision applications.
Let's take a closer look…
Backside Illumination The "S" in Pregius S stands for "stacked," but it could just as easily stand for "sensitivity & speed" because that's what the stacked design of the Pregius S pixels provides.
Traditional image sensors have wiring and circuitry on the layer above the photodiodes. This results in a somewhat narrow angle of incidence for the incoming light to strike the photodiodes and fill the "pixel wells."
In the Pregius S structure, the wiring and circuitry layers are stacked below the photodiodes. You could also think of this as flipping the sensor over and capturing the light via the back of the photodiodes – hence the term "backside illumination" (BSI). Getting the wiring out of the way and moving the photodiodes closer to the microlenses creates a much wider angle of incidence for capturing incoming light. As a result, BSI pixels are more sensitive than traditional "frontside illuminated" pixels.
Sony has leveraged this BSI design in several ways. First, they've created a small pixel size – 2.74 µm per side – but because of BSI it has the sensitivity you'd expect from a much larger pixel. They also found a way to reduce read noise to a level lower than any of their previous Pregius generations (2.1 e-). This creates dramatic improvements in signal-to-noise ratios, saturation capacity, and overall image quality. Lastly, because they didn't have to create openings in the wiring area for the light to reach the pixels, Sony has used the extra space to design a new, faster sensor interface (SLVS-EC), providing higher frame rates and extra signal processing capabilities.
The result is a pixel design capable of powering a new generation of JAI machine vision cameras for use in a wide range of vision systems.
Global Shutters Sony's BSI technology has previously only been available in sensors with rolling shutters. While inexpensive rolling shutter sensors help reduce camera costs, the resulting cameras produce "skewed" images of moving objects making them unsuitable for many types of machine vision applications. The Pregius S series combines BSI technology with global shutters, enabling cameras that can capture crisp, undistorted images in all types of applications, including those where significant movement is involved.
High resolution, small optical circles The 2.74 µm pixel size means that cameras with Pregius S sensors can pack higher resolutions into smaller physical sensor sizes – meaning more pixels per square-millimeter of silicon. This means the sensors carry a lower cost-per-megapixel, and this savings can get passed along to the camera cost.
This gives system designers a couple of options:
They can save by reducing the number of cameras needed per system. For example, an inspection system that needs two 12-megapixel cameras can possibly get by with only one 24-megapixel camera.
Or, they can keep the same number of cameras, but double the resolution of their system, increasing the value of their system by enabling it to identify smaller defects or inspect larger fields of view.
Lastly, the small pixels mean more resolution without having to switch to F-mount, M42-mount, or other large, heavy, and expensive lens formats. JAI's Pregius S-based cameras feature resolutions from 5 to 24 megapixels with optical formats from 1/1.8" to 1.2". This means even the highest resolution models can utilize C-mount lenses, saving on size, weight, and costs.
A choice of aspect ratios Pregius S sensors have been created with several different aspect ratios, resulting in cameras that can efficiently apply the available resolution to a range of application types. Included are:
Traditional 4:3 and 5:4 ratios popular for many machine vision applications because they pack the resolution into the highest quality region of typical C-mount lens circles.
1:1 ratio used in microscopy applications (medical & industrial) and metrology.
A 16:9 ratio suited to HD screens like in traffic & transportation, sports, virtual reality, and some web/surface inspection applications.
Additional on-sensor functions As noted above, the extra circuitry area made possible by the Pregius S "stacked" architecture has enabled the addition of several more on-sensor signal processing capabilities, which have been leveraged into JAI cameras. Two examples of this are:
Gradation compression – For scenes containing a few very bright areas, this function lets the camera capture a full 12-bits of shading variations but compresses the output into 8-bit images to increase frame rates and decrease file sizes. Users can define how much compression to apply and where on the 12-bit scale it should begin.
Floating diffusion binning – In addition to 1x2, 2x1, and 2x2 binning performed on the camera's FPGA, JAI's monochrome Pregius S cameras offer 2x2 binning performed directly on the sensor. This "FD2x2 Binning" mode utilizes a readout strategy that combines the readout of four pixels through a single floating diffusion node. This not only lowers the readout noise for higher dynamic range and image quality, it also produces significant increases in frame rate over the FPGA-based binning modes.
A decimation function for color models, multi-ROI capability, and several other camera features are also enabled by the on-sensor capabilities of the Pregius S sensors.
Put Pregius S technology to work for you Sony Pregius S sensor technology is available in selected models within JAI's Go-X Series of small, affordable, and reliable machine vision cameras. A total of 24 Pregius S-based models are currently available including 12 with CoaXPress interfaces and 12 with GigE Vision (1000BASE-T) interfaces. 12 more models with GigE Vision 5GBASE-T interfaces will be available in the near future.