Radar Detection With Sliding Spectrogram Windows for Point Density
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for improving point cloud resolution in laser radars, such as increasing the number of lasers, lead to increased costs, size, and power consumption without effectively enhancing the signal-to-noise ratio.
Innovation Solution
Transforming the beat frequency signal into a two-dimensional spectrogram and intercepting measurement units (MUs) based on a time domain sliding step, where the time domain length is less than the frequency modulation period, allowing for higher point output rates without compromising the signal-to-noise ratio or increasing costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the number of lasers is increased to improve point cloud resolution, then the point cloud density is improved, but the cost, size, and power consumption significantly increase
Solution Approach 1:
The patent segments the frequency modulation period into multiple measurement units (MUs) by applying short-time Fourier transform with overlapping time windows. This allows multiple measurements to be extracted from a single frequency modulation period, effectively increasing point cloud density without adding hardware lasers. The segmentation of time domain signals enables parallel processing of multiple measurement instances.
Solution Approach 2:
The patent transitions from spatial dimension (adding more lasers) to time-frequency dimension (processing signals in time and frequency domains). By using short-time Fourier transform and analyzing signals in the time-frequency plane, the system extracts multiple measurement units from temporal variations within a single frequency modulation period, achieving increased point cloud density through temporal rather than spatial multiplication.
2Reliability
If the measurement time domain length is increased to improve signal-to-noise ratio, then the signal-to-noise ratio is improved, but the point output rate decreases
Solution Approach 1:
The patent applies preliminary action by using short-time Fourier transform to pre-process the beat frequency signal and identify multiple measurement units within a single frequency modulation period before final detection. This pre-processing enables the system to extract multiple valid measurements from temporal variations, allowing shorter effective measurement times while maintaining sufficient signal-to-noise ratio through cumulative processing of multiple MUs.
Solution Approach 2:
The patent implements continuous useful action by using overlapping time windows in the short-time Fourier transform process. The overlapping ensures that signal energy is continuously captured across different time segments, and multiple measurement units are extracted continuously from the same frequency modulation period, maximizing the utilization of available signal energy while increasing output rate.
3Productivity
If traditional FFT processing is used on beat frequency signal, then the processing is simple, but the point cloud density is low
Solution Approach 1:
The patent introduces dynamics by transitioning from static FFT processing to dynamic short-time Fourier transform with sliding time windows. The processing adapts to temporal variations in the beat frequency signal by continuously analyzing different time segments with overlapping windows, enabling extraction of multiple measurement units that reflect dynamic changes within the frequency modulation period.
Solution Approach 2:
The patent introduces an intermediary processing stage between signal acquisition and final detection. The short-time Fourier transform acts as an intermediary that transforms the time-domain beat frequency signal into a time-frequency representation, enabling intermediate extraction of multiple measurement units. This intermediary processing layer bridges the gap between simple FFT and complex multi-laser systems.
Data Source
AI summary
A radar detection method includes transforming a beat frequency signal of a radar into a two-dimensional spectrogram; intercepting, based on a time domain sliding step, a plurality of measurement units (MUs) whose time domain lengths are equal to a frequency modulation period of the radar from the two-dimensional spectrogram, where a length of the time domain sliding step is less than the frequency modulation period of the radar; and determining a radar detection result based on each of the plurality of MUs.