Depth Map Pixel Validation for Precise Distance Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing distance measuring devices, particularly those integrated into mobile terminals like smartphones, often produce false distance measurement results, necessitating improved methods for detecting and correcting such inaccuracies to enhance measurement precision.
Innovation Solution
The proposed solution involves a two-step process using determination units to assess differences in depth values and confidence or reflectance between adjacent pixels in a depth map, identifying defective pixels by comparing these values against thresholds, and employing a machine-learned learning model to correct these pixels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single threshold comparison method is used to detect defective pixels, then the detection process is simple, but the detection precision is insufficient leading to false positives
Solution Approach 1:
The patent divides the defective pixel detection process into two distinct stages: first comparing depth values between adjacent pixels, and then comparing confidence values or reflectance values. This segmentation allows each comparison to focus on specific characteristics, improving overall detection precision while maintaining a structured and manageable process complexity
Solution Approach 2:
The patent performs more comparisons than a single threshold method would require by conducting both depth value comparison and confidence/reflectance value comparison. This excessive action ensures that false positives are reduced by requiring multiple conditions to be met, thereby improving detection precision at the cost of additional computational steps
2Volume of moving object
If distance measuring devices are miniaturized for mobile terminals, then the device size is reduced, but measurement accuracy deteriorates due to false distance measurement results
Solution Approach 1:
The patent uses confidence values or reflectance values as auxiliary information (copies) that parallel the depth measurement data. By comparing these auxiliary copies with the primary depth measurements, the system can verify the validity of distance measurements without requiring additional physical sensors, thus maintaining miniaturization while improving accuracy
Solution Approach 2:
The patent implements a feedback mechanism where confidence values or reflectance values provide verification information back to the depth measurement process. This feedback loop allows the system to identify and correct false distance measurements by detecting inconsistencies between the primary measurement and the auxiliary verification data
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 effectively detects and corrects false distance measurements, enhancing the accuracy of distance measurement results by identifying and rectifying defective pixels, thereby improving the overall precision of the measurement process.
Implementation Method 1
a light receiving unit that receives the reflected light and measures a distance to the object
Implementation Method 2
an indirect TOF method in which a light emitting unit emits modulated light to an object
Data Source
AI summary
The present technology relates to a distance measuring device, a distance measuring method, a program, an electronic apparatus, a learning model generating method, a manufacturing method, and a depth map generating method that are designed to enable distance measurement with higher precision.The distance measuring device includes: a first determination unit that determines whether or not the difference in depth value between a first pixel in a depth map and a second pixel adjacent to the first pixel is larger than a first threshold; and a second determination unit that determines whether or not the difference in confidence between the first pixel and the second pixel is larger than a second threshold, in a case where the first determination unit determines that the difference in distance between the first pixel and the second pixel is larger than the first threshold. In a case where the second determination unit determines that the difference in confidence between the first pixel and the second pixel is larger than the second threshold, the first pixel is confirmed to be a defective pixel. The present technology can be applied to a distance measuring device, for example.


