Also called digital image stabilization, electronic image stabilization (EIS) is a digital technique that helps capture sharper pictures. In other words, EIS minimizes blurriness in pictures and shakiness in videos induced by unsteady camera handling or movement. EIS tries to do what optical image stabilization (OIS) achieves, but it does that electronically or through software. This means the end results with EIS are generally not on par with OIS.
When there’s camera shake, the camera sensors detect the movement. EIS counters that by moving the image slightly so that the picture doesn’t move out of the frame. Contrary to OIS, EIS solves the issue of blur and shakiness at the programming level with the help of a charge-coupled device (CCD), the camera’s light-sensing chip. The chip detects camera movement using its sensors and compensates for the same after marginally moving the picture and ensuring its positioning on the screen doesn’t change. All of this happens after the picture has been captured or after the conversion of the optical signal into a digital signal, which happens within a fraction of seconds. With OIS, all the adjustments happen before the image is captured.
During the process, the EIS technology zooms in to the image slightly to decrease image movement. The resulting inflated image makes it easier for the system to detect shake-induced movement. However, such increase in image size can decrease the resolution of the picture, which is probably why most cameras (smartphone or conventional cameras) with electronic image stabilization have a higher resolution.
There is another electronic stabilization method wherein the system uses a much larger CCD, in which the video or image takes up only 90 percent of the CCD’s overall footprint. The remaining space is used for moving the picture. In other words, if the shot is steady, the chip would have the image in the center. However, if the camera is shaking, the chip would make use of its empty real estate to reposition the image.
Similar to OIS, EIS cannot do much if the blurriness was from the subject’s end or when the camera is shaken vehemently. Also, at the algorithm front, EIS may have issues differentiating between intentional tilts and haphazard camera movement. This is more an issue when zooming in on subjects because that could result in the subject taking too much frame space and the movement becoming more noticeable.
Comparing EIS and OIS
EIS, compared to OIS, is cheaper and more feasible to be implemented on a wider scale. For instance, it can be used on surveillance cameras found affixed to traffic signal posts to cut out the jitteriness caused by passing traffic or winds, resulting in a much smoother live capture. OIS can probably do a better job, but it’s expensive and is not designed to be used in such locations since OIS has moving components, which makes it a bit fragile. EIS is also much quicker as there’s no physical movement involved and everything happens electronically. Moreover, EIS video file sizes are smaller than the size of OIS-enabled footage.