Aerial vector gravity measurement not only obtains the vertical component of gravity disturbances, but also the horizontal component, i.e., the deflection of the vertical. The strapdown airborne gravimeter is the main instrument for measuring gravity disturbance vectors, primarily composed of SINS and GNSS receivers, with SINS being the core component. SINS consists of a three-axis gyroscope and accelerometer. A 1” horizontal attitude error caused by gyroscope error will bring a measurement error of 4.75 mGal to the horizontal component of gravity disturbance, making gyroscope error a decisive factor restricting the development of aerial vector gravity measurement. Therefore, many scholars focus their research on how to separate the gravity disturbance component from the accelerometer observation data that is greatly affected by attitude error. This chapter introduces the SINS coordinate transformation, mechanical arrangement, error equation, and Kalman filtering, with a focus on the initial alignment algorithm based on Kalman filtering, filtering calculation under local horizontal coordinates, and filtering calculation under inertial coordinates, as well as a series of error correction methods carried out later to further separate attitude angle error, acceleration error calculated from GNSS observations, accelerometer error, etc.