I. Camera Shake
When a smartphone is used to photograph a child running through a park, camera shake can turn a lively moment into a blurred image. During aerial photography, severe shake can make footage captured by a drone unstable and diminish the visual impact of the landscape. In security surveillance, blurred footage at a critical moment may obscure important information. These apparently different problems share the same underlying cause: camera shake.

Camera shake is the unintended displacement or vibration of a camera or other imaging device during image capture, usually caused by external disturbances. This instability can cause still images to exhibit motion blur and loss of detail, while video recordings may show vertical or horizontal jitter and misalignment between consecutive frames. As a result, both the viewing experience and the effectiveness of visual information transmission may be significantly impaired.

II. Image Stabilization Technologies

Image stabilization reduces the effects of camera motion through optical, mechanical, or computational methods. The two main approaches are optical image stabilization (OIS) and electronic image stabilization (EIS).

II.1 Optical Image Stabilization (OIS)

OIS uses motion sensors, typically gyroscopes, to detect camera movement. An actuator then shifts the lens group or image sensor to compensate for changes in the optical axis.

Because the compensation is performed optically before image capture, OIS generally does not require image cropping and can preserve the full field of view. It is particularly useful for telephoto photography and low-light capture, where camera movement is more likely to cause visible blur.

II.2 Electronic Image Stabilization (EIS)

EIS uses software algorithms to estimate camera motion and align successive frames or image regions. It reserves additional pixels around the image area and crops or transforms the frames to compensate for the estimated movement.

EIS requires no dedicated optical compensation mechanism and can reduce hardware complexity. It is widely used for video recording, and some devices combine image-based motion estimation with gyroscope data to improve stabilization performance.

Its main limitation is the reduction of the field of view caused by cropping. The amount of cropping varies by device and operating mode and is typically about 5%–15%. Excessive cropping or image processing may also reduce image quality.

Although OIS and EIS can substantially reduce the effects of camera shake, neither can eliminate unwanted motion completely. Their effectiveness depends on factors such as the stabilization method, the magnitude and frequency of the movement, exposure time, and the device’s operating mode. Since stabilization performance varies between devices and shooting conditions, it should be measured under controlled and representative scenarios.

III. Image Stabilization Test Method

Video stabilization testing reveals two immediately visible artifacts: overall image jitter and motion blur. The objective of stabilization measurement is to evaluate video stability under controlled mechanical vibration by quantifying two phenomena: motion blur caused by residual shake and geometric distortion introduced by the stabilization process.

III.1 Test Equipment:
Vertical multi-color temperature fill light source (LS-CCXL-2S06-IR), test chart (CP316 Reflective Gray Dead Leaves Chart), 2-axis stabilization kit (IS-Shake02)

The IS-Shaker02 image stabilization test system is a high-precision test equipment independently developed by Yanding, specifically designed for evaluating camera image stabilization performance. It can simulate real-world shake scenarios through high-precision dual-axis vibration, and is applied to the image stabilization performance evaluation of devices such as smartphones, smart glasses, security cameras, and DSLR cameras.

III.2 Test Procedures:

1. Environment Setup:
Conduct the test in a darkroom with an indoor illuminance of ≤2 lux; the illuminance uniformity on the surface of the test chart should be ≥90%;
Lighting environment: 6500K, 3000 lux; 6500K, 800 lux; 6500K, 30 lux;
Shake platform set frequencies and angles: 2Hz, 4°; 4Hz, 2.5°; 6Hz, 1.5°; 8Hz, 1°; 10Hz, 0.8°; 12Hz, 0.15°; 15Hz, 0.15°.

2. Sample Acquisition
The DUT must be fixed to the 2-axis stabilization platform so that the geometric center of the DUT coincides as much as possible with the rotation center of the platform, ensuring that its optical axis is as perpendicular as possible to the plane of the test chart. The test chart should occupy about one-third of the image height. After the image stabilizes and focus is clear, start the 2-axis stabilization platform and record video for a duration of no less than 15 seconds.

3. RIQA Analysis
Open the RIQA software, click on the chart directory, and select Shaking in the Camera video module. Click “+Add” to add the video to be analyzed, and set the frame interval. Enter 1 if every frame needs to be analyzed, or enter an appropriate frame interval, such as 10, to reduce the analysis time.

Click the “Start” button, and the software will automatically analyze and output the data metric results of the recorded shake video. Click the “Generate Report” button to generate a detailed test report containing key metrics, chart comparisons, and clear conclusions with one click.

4. Result Interpretation

The data metric results are described as follows: