Rolling Shutter Correction
Rolling shutter correction is a related family of techniques for removing image warping produced by intra-frame camera motion. High-end cameras use CCD (Charged Coupled Device) sensors, which have a global shutter (GS). In a GS camera all pixels on the CCD sensor are read out and reset simultaneously. Therefore all pixels collect light during the same time interval. Consequently, camera motion during the exposure results in some amount of image blur on these devices. In contrast, most digital cameras typically make use of CMOS (Complementary Metal Oxide Sensors). These sensors use a rolling shutter, where image rows are read out and reset sequentially. The advantage of this approach is that it requires less circuitry compared to CCD sensors. This makes CMOS sensors less expensive to manufacture. For that reason, CMOS sensors are frequently used in camera phones, digital cameras, and consumer to prosumer camcorders.
Sequential readout, however, means that each row is exposed during a slightly different time window. The rolling shutter effect is hence essentially a positional error in the video, which occurs in moving footage when the frame is scanned line by line. If the image is static, the footage is correct because all points of the frame are exposed at their correct positions. However, if the camera is panning quickly (eg: video shot from a moving vehicle), or the object is moving rapidly (eg: plane propeller blades) the rolling shutter effect will result in digital videos that wobble, skew (diagonal distortion), smear or have only partial exposure. This is especially apparent when recording fast-moving subjects.
Also, if the camcorder is subject to heavy shaking, as may occur when capturing with a point of view camera (e.g. a GoPro camera mounted on a mountain bike heading downhill past a swathe of pine trees), strong distortion of the entire image may result, severely impairing the quality of the video footage. The normally vertical and straight pine trees will appear like parallelograms, slanting diagonally.
This is a very well know problem and many techniques have been implemented to achieve the rolling shutter fix.
Here are a few popular techniques.
MEM gyroscopes within a cellphone (e.g. iPhones and Android Smartphones) or camera can be used to measure camera rotations. These measurements are then used in conjunction with compensatory algorithms to achieve rolling shutter correction in the recorded video image.
Using a theoretical analysis and a mathematical modeling of the rolling shutter distortion, DSPs are sometimes used in devices to compensate this effect. A polynomial curve fitting is developed to approximate the horizontal and vertical velocities. To increase usable accuracy, local motion estimation is performed in each small region. This method is computationally expensive and only works in two dimensions.
Best Rolling Shutter Fix
Turbo Video Stabilizer employs a full three dimensional modeling to measure all the motions of the camera in the X, Y and Z axes, plus the rotational moments: roll, yaw and pitch. This 6-axes model is then used to stabilize videos and also fix any rolling shutter artifacts.
muvee Turbo Video Stabilizer is the best software for post-capture rolling shutter fix for point of view videos like those captured with GoPro, Contour and Sony Action Cam, all of which uses CMOS sensors. Videos shot with the iPhone and even prosumer SLRs are also susceptible to the rolling shutter problem.
The Rolling shutter distortion tends to be undetectable or hardly noticeable in footage before digital video stabilization. But after a video is successfully stabilized in post-production, the rolling shutter effect remains and may slightly or severely impair the quality of the footage. For this reason, muvee Turbo Video Stablizer is also equipped with a powerful rolling shutter fix algorithm to combat this unwanted effect. Afterall, there is no point in a video stabilization fix if the resulting video is degraded by other digital video artifacts.