Modeling and correction of depth error in indirect time-of-flight cameras

Abstract The ranging performance of time-of-flight (ToF) cameras is mainly influenced by the signal harmonics, measurement noise and number of repeated integrations. However, current depth error correction methods lack a synthetic-factor model to elucidate the error generation mechanism deducing the image quality degradation of the ToF camera. To overcome the shortcomings of the current methods, a depth error model, combining the wiggling error, temporal error and integration time-related error, was proposed to achieve single-pixel high-precision distance correction based on the charge transfer principle. A vertical center-pixel depth error calibration method was used to obtain the real distance deviation and validate the depth error model under various distances and integration times. Based on the probability distribution of all-pixel depth errors, a segmented pixel region correction method was advanced to suppress the randomness error of each pixel. Based on a commercial ToF camera, a series of depth error correction experiments was carried out. The experimental results show that the distance deviations of all-pixel imaging are mainly centered on 0.26 ± 3 mm in the range of 2.6 m, and its integration time is 4 ms. The proposed depth error model provides an effective way to enhance the ToF camera ranging accuracy, serving a wider audience of researchers and industrial practitioners.