Depth Map Correction Using Grayscale Histogram Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods face challenges in accurately estimating depth information due to external factors like lighting and environment, particularly in overexposed or underexposed images, leading to inaccurate depth maps.
Innovation Solution
An image processing method that utilizes a grayscale histogram to determine if an image requires depth correction, and through a series of steps, corrects the depth map by identifying high-luminance targets and adjusting depth values based on threshold conditions, improving the accuracy and efficiency of depth estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If depth estimation is performed on images with strong light sources, then processing speed is maintained, but depth estimation accuracy deteriorates
Solution Approach 1:
The patent applies preliminary action by analyzing the grayscale histogram before depth estimation to detect overexposed or underexposed conditions. This preliminary detection allows the system to identify images with strong light sources that would compromise depth estimation accuracy, enabling preprocessing or alternative processing strategies to be applied before the main depth estimation operation.
Solution Approach 2:
The patent implements feedback by using the grayscale histogram analysis results to guide the depth estimation process. The histogram characteristics provide feedback about image quality and lighting conditions, which then informs whether to proceed with standard depth estimation or apply correction measures, creating a closed-loop system that improves accuracy based on detected conditions.
2Measurement precision
If depth correction processing is applied to all images, then depth estimation accuracy improves, but processing time increases
Solution Approach 1:
The patent applies partial action by performing depth correction processing only on images that are identified as overexposed or underexposed through grayscale histogram analysis. Instead of applying correction to all images, the system selectively processes only those images that exhibit characteristics indicating they would benefit from correction, thereby reducing overall processing time while maintaining accuracy where needed.
Solution Approach 2:
The patent uses parameter changes by analyzing the grayscale histogram parameters (such as the distribution of pixel intensities) to determine whether an image requires depth correction. By monitoring these parameters, the system can dynamically adjust its processing approach, applying correction only when the histogram parameters indicate degraded image quality that would affect depth estimation accuracy.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
An image processing method according to an embodiment of the present invention includes acquiring consecutive frame images, generating a depth map and a grayscale histogram of a current frame image of the consecutive frame images, outputting, in a case where the grayscale histogram of the current frame image does not satisfy a first condition, the depth map of the current frame image, determining, in a case where the grayscale histogram of the current frame image satisfies the first condition, whether the grayscale histogram of the current frame image satisfies a second condition, and correcting, in a case where the grayscale histogram of the current frame image satisfies the second condition, the depth map of the current frame image, and outputting the corrected depth map of the current frame image.