Search

This new methodology selectively scans an area of interest and effectively pre-compresses the image data. Instead of using LiDAR resources to gather redundant data, only the necessary data is gathered and the redundancy can be used to fill in up-sampled data using intelligent completion algorithms. The system utilizes a unique LiDAR system to collect a pattern of specific points across a given area by modulating the incoming light, creating a pattern that can be decoded computationally to reconstruct a scene. By designing specific coding patterns, the system can strategically skip certain measurements during the scanning process to create an under-sampled image area. The system reconstructs the under-sampled area to recreate an accurate representation of the original object or area being scanned. As a result, redundant data is prevented from being collected by reducing the number of required measurements and data condensed in post-collection to reduce power consumption. By selectively skipping certain pixels during the scan and using sophisticated recovery algorithms to reconstruct the omitted information, the system makes more efficient use of the available photons, thereby enhancing overall data collection. This technology represents a significant advancement in LiDAR systems, offering a more useful method for data collection and processing and addresses the challenges of power consumption and data redundancy, allowing for more sustainable and effective remote sensing applications. This technology can offer advantages in applications such as mapping for construction, surveying, forestry, or farming as well as computer vision for vehicles or robotics.