面向极低照度的SPAD主动成像数据压缩方法
By performing wavelet transform and Gaussian filtering on the histogram data of the SPAD array, the problem of poor depth reconstruction performance of the SPAD array under extremely low illumination was solved, and the data rate was reduced and noise was suppressed, thereby improving the accuracy and efficiency of depth reconstruction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIDIAN UNIV
- Filing Date
- 2025-07-04
- Publication Date
- 2026-07-17
AI Technical Summary
SPAD arrays exhibit poor depth reconstruction performance under extremely low illumination conditions. They also have high data rates and are susceptible to ambient light noise, resulting in extremely poor depth reconstruction performance.
Wavelet transform is used to decompose the histogram data of a single pixel in the time-frequency domain. Combined with nonmaximum suppression of adjacent pixels and adaptive Gaussian filtering, the data is smoothed by Gaussian filtering. The multi-resolution characteristics and spatial correlation of wavelet transform are used for data compression.
At low photon count levels, data bandwidth compression of two orders of magnitude was achieved, while depth reconstruction performance was improved, noise impact was reduced, and the accuracy and efficiency of depth reconstruction were enhanced.
Smart Images
Figure CN121169985B_ABST