Image sensors, the eyes of electronic products, can provide the machine vision for various applications. Though it is not a new device anymore, the related technologies are still advancing and lead to the performance with a higher resolution, better image quality and lower power consumption. In this article, we’ll introduce the latest development trend of the image sensor.
Global shutter that can prevent the jello effect
The technology of the image sensor has been developing for many years. At the early stage, the development was mainly focused on CCD technology that is applied to different video cameras. With the trend of digitalization and as portable devices focus more and more on the usage time, CMOS technology, with features like power saving and easy integration has gradually become the mainstream and is popularly adopted by electronic products including digital cameras, video cameras and mobile phones.
There are no merits and demerits of the image quality brought by these two image transmission technologies. However, under the same condition, when comparing with CCD sensors, CMOS ones require fewer components, therefore leading to less power consumption while achieving high data throughput. Since CMOS sensors of the digital camera can be directly assembled on the mainboard circuit, the signal transmission distance is shorter than CCD ones. RC and stray delays can thus be reduced. Moreover, the lower manufacturing cost also promotes CMOS technology’s domination over the current market.
For past CMOS sensors, the introduction of the rolling shutter would generate the undesired jello effect when shooting fast-moving objects. This phenomenon happens because CMOS sensors need to run progressive scanning when recording optoelectronic signals in the style of point-to-point image recording from left to right and up to down, just like lowering a curtain. It will inevitably result in the time difference when reading signals from the first row to the last one. When shooting fast-moving scenes, the image will tilt and deform because of the time difference and it consequently lead to the well-known jello effect.
How can we solve this jello effect? The effect can be solved by not recording data by point-to-point or line-to-line; however, this method will impact CMOS’s performance on high speed, high efficiency and low power. Meanwhile, with the improvement of processor and memory technologies, the integration of both is no longer a dream and the global shutter is thus created.
In this paragraph, we’ll take AR0134 image sensor from ON Semiconductor for Oculus Rift VR device as an example. This industrial grade 1/3-inch MP sensor with global shutter can be used for applications such as barcode scanning, 3D scanning, position tracking, virtual reality (VR), augmented reality (AR), biometrics, and machine vision. With a pixel size of 3.75µm and the resolution up to 1.2Mp (1280x960), this device can provide superior low-light performance, linear capture technology, supports HD video output (720p 60 fps), video/single frame model, flexible row-skip modes and parallel and serial output and comes with auto exposure control, auto black level calibration and statistics engine. This 3rd generation global shutter technology can provide the global shutter pixel with the highest performance, the responsivity up to 6.1V/lux-sec and the frame rate up to 45 fps with full resolution. It also supports the 720p 60 fps video application and can work with external LED or flash. Meanwhile, context switching can support applications with different imaging requirements and operate effectively through a single source.
ON Semiconductor focuses on significantly improving pixel performance and sets the foundation for superior image quality and performance of sensors. AR0134 sensor integrates the high-performance global shutter technology of ON Semiconductor and captures as well as integrates fast-moving images to HD devices of 1/3-inch optical format. The superior low-light performance of 3.75µm global shutter pixel can capture movement and prevent the jello effect usually generated by rolling shutter pixel from occurring.
Fueled by VR and AR applications, the pace of image sensor market development is accelerating. Technological progress enhances shooting quality, power consumption and performance of the image sensor, to achieve a sharper and more extensively applied eye of the machine as well as a rapidly developing market.
