Unlock instant, AI-driven research and patent intelligence for your innovation.

Long-distance laser radar depth estimation method based on intensity guidance

A lidar and depth estimation technology, applied in computing, image enhancement, image analysis, etc., can solve problems such as small memory requirements, depth map errors, etc., to achieve the effect of accelerating computing speed, improving robustness, and reducing data size

Pending Publication Date: 2022-01-28
NANJING UNIV OF SCI & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide an intensity-guided long-distance laser radar depth estimation method to solve the problem that the traditional method has a large error in the depth map estimated from the long-distance laser radar 3D point cloud with a large amount of data and is extremely sparse. , thus estimating a more accurate depth image with less memory requirement and computational complexity

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Long-distance laser radar depth estimation method based on intensity guidance
  • Long-distance laser radar depth estimation method based on intensity guidance
  • Long-distance laser radar depth estimation method based on intensity guidance

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] Such as figure 1 As shown, it is a schematic diagram of a target scene used to verify an intensity-guided long-distance lidar depth estimation method of the present invention. The target scene is about 1.4km away from the lidar system, and the peak of the histogram is the center, and the data with a total length of 180m before and after is retained, so as to obtain the original 3D point cloud with a size of 200×200×3000.

[0045] Such as figure 2 Shown is the target reference depth image and 3D point cloud image obtained when the sampling integration time is 50ms. figure 2 (a) depth map, figure 2 (b) is a 3D point cloud image.

[0046] Such as image 3 Shown is the target depth map and 3D point cloud map acquired by the traditional imaging method based on the maximum likelihood estimation algorithm when the sampling integration time is 10ms. image 3 (a) depth map, image 3 (b) is a 3D point cloud image.

[0047] Such as Figure 4 Shown is the target depth ma...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a long-distance laser radar depth estimation method based on intensity guidance. The method comprises the following steps: enabling space-time correlations and intensity information of an echo signal to be fully utilized in data preprocessing, and alternately carrying out multi-scale superpixel building and rapid time domain windowing on laser radar three-dimensional point cloud data; based on a Poisson distribution model of the preprocessed laser radar three-dimensional point cloud data, adopting an optimization framework to establish a depth estimation cost function in which a data fidelity term and a regular term introduced by the intensity information are combined; and using an alternating direction method of multipliers to estimate a depth image from the cost function. Aiming at the problems that three-dimensional point clouds obtained by long-distance laser radar are sparse and the amount of data is large, the invention estimates a depth image with a smaller memory requirement and lower calculation complexity, effectively improves the quality of the depth image, and improves the robustness to a low photon level.

Description

technical field [0001] The invention belongs to laser radar depth estimation technology, in particular to a long-distance laser radar depth estimation method based on intensity guidance. Background technique [0002] Long-distance 3D imaging has been widely used in terrain surveying, underwater exploration, 3D remote sensing and other fields. The lidar system uses an avalanche photodiode (GM-APD) operating in Geiger mode, which has single-photon sensitivity and picosecond-level temporal resolution. The photon technical statistical histogram of each pixel is obtained by time-correlated single photon counting (TCSPC) technology, in which the counted position and quantity correspond to the depth and intensity information respectively, so as to reconstruct the 3D scene. [0003] Long-distance photon counting lidar continues to develop and can work at longer distances, larger fields of view, and higher resolution requirements, which means that the obtained lidar 3D data size is ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/521G06T5/00G06T5/40
CPCG06T7/521G06T5/40G06T2207/10028G06T2207/20182G06T5/70
Inventor 何伟基邬淼张闻文陈钱邹燕
Owner NANJING UNIV OF SCI & TECH