Systems and methods for translational motion correction

Video information generated by an image sensor for a capture period and motion information of the image sensor during the capture period may be obtained. The video information may define images of a video based on light received within a field of view of an image sensor during the capture period. Relative positions of the image sensor between different moments within the capture period may be estimated based on the motion information. Depth information for a portion of an environment of the image sensor may be determined based on the video information generated at the different moments and the relative positions of the image sensor between the different moments, or based on stereo depth mapping. Translational motion correction may be applied to one or more of the images based on the depth information and the relative positions of the image sensor to warp the image(s) and stabilize the video.