Synchronous Event-Driven Readout for Direct Time-of-Flight Depth Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional direct time-of-flight depth cameras face inefficiencies in data readout, leading to low-quality depth reconstruction due to fixed frame rate reading, which results in high activity pixels being read less often and older photon detections being identified multiple times.
Innovation Solution
The implementation of a depth camera assembly with an imaging device configured to capture images using macropixels, where each macropixel is coupled to readout buses via row and column access control signals, allowing for increased readout speed and accurate evaluation of photon detections by grouping pixels into macropixels and using time-to-digital converters to generate digital timestamps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fixed frame rate is used to read detector data, then the readout process is simple and regular, but high activity pixels are read less often and depth reconstruction quality deteriorates
Solution Approach 1:
The patent implements dynamic readout scheduling where pixels are read at different rates based on their activity levels. High activity pixels are read more frequently while low activity pixels are read less frequently, replacing the static fixed frame rate approach with an adaptive dynamic system that optimizes both readout simplicity and depth reconstruction quality
2Stability of the object's composition
If a fixed frame rate is used to read detector data, then the readout timing is regular and predictable, but older photon detections are identified multiple times
Solution Approach 1:
The patent employs feedback mechanisms where the system monitors photon detection activity and adjusts readout scheduling accordingly. This feedback loop prevents redundant reading of older photon detections by dynamically adapting the readout timing based on actual pixel activity, thereby eliminating information loss while maintaining operational stability
3Device complexity
If all pixels are read at the same rate, then the readout process is uniform and simple, but data from high activity pixels is insufficient and data from low activity pixels is redundant
Solution Approach 1:
The patent applies local quality differentiation by assigning different readout rates to different pixels based on their individual activity levels. Each pixel or pixel group receives customized readout scheduling that matches its actual photon detection characteristics, optimizing data acquisition efficiency while managing system complexity through localized adaptation rather than global uniformity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the speed and accuracy of data retrieval from the detector, improving depth reconstruction by selectively accessing and processing data from high and low activity pixels, reducing noise and improving signal-to-noise ratio.
Implementation Method 1
single-photon detectors are used, such as single-photon avalanche diodes (SPADs)
Implementation Method 2
time-to-digital converters to generate digital timestamps
Implementation Method 3
measure a roundtrip travel time of photons generated by multiple short pulses of light from an illumination source
Implementation Method 4
Direct time-of-flight allows multiple events (e.g., detections of photons) to be acquired in a histogram through a process called time-correlated single-photon counting (TCSPC)
Data Source
AI summary
Each pixel of a plurality of pixels captures a portion of outgoing light illuminating a local area reflected from the one or more objects in the local area. A flag is generated via a flag determination logic circuit coupled to each pixel of the plurality of pixels. The flag indicates whether the reflected light was captured within a threshold amount of time from a current time. A time to digital converter is enabled via a flag determination logic circuit. The time to digital converter is associated with each pixel of the plurality of pixels to generate a digital representation of time when each pixel of the plurality of pixels captured the reflected light. Depth information is determined for the one or more objects in the local area based in part on the generated flag and the digital representation of time.