Adaptive Gray-Level Correction for Low-Contrast Object Counting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques for estimating the number of specific objects in images, such as human bodies and vehicles, suffer from reduced accuracy in low contrast conditions, especially in fog, mist, or through glass, and are not optimized for enhancing the estimation of specific object counts.
Innovation Solution
An image processing apparatus and method that employs gray-level correction using weighted histograms and correction curves to emphasize contrast in low luminance regions, combined with a number estimation unit to select the largest estimated number, and a system that dynamically adjusts correction curves based on object density and motion analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If conventional gray-level correction techniques are applied to improve image quality, then image contrast is enhanced, but estimation accuracy of specific objects does not improve because the correction is not optimized for object counting
Solution Approach 1:
The patent applies different gray-level correction strategies to different regions of the image based on local characteristics. The system divides the image into multiple regions and performs histogram analysis separately for each region, applying correction amounts tailored to local luminance distributions rather than uniform global correction.
Solution Approach 2:
The patent dynamically adjusts correction parameters based on image analysis. The correction amount is determined by analyzing the histogram distribution and calculating optimal correction values that enhance object detectability. Multiple correction amounts are computed and the most appropriate one is selected based on the actual image content.
2Device complexity
If a single tone-correction method is applied to all images, then processing is simple, but accuracy decreases in low luminance regions such as fog or mist conditions
Solution Approach 1:
The patent implements dynamic correction by computing multiple correction amounts based on different histogram analysis methods and selecting the most appropriate correction for each image. The system adapts the correction strategy based on the actual luminance distribution and object characteristics in the input image, rather than applying a fixed correction method.
Solution Approach 2:
The patent segments the correction process into multiple independent correction amounts, each derived from different histogram analysis approaches. This allows the system to evaluate multiple correction strategies and select the optimal one, effectively segmenting the solution space to handle diverse imaging conditions.
3Measurement precision
If feature points are calculated from images for object estimation, then object counting is enabled, but accuracy decreases in low contrast states such as fog, mist, or glass conditions
Solution Approach 1:
The patent performs preliminary gray-level correction on images before conducting feature point calculation and object estimation. By pre-enhancing the image contrast through optimized correction, the subsequent object detection and counting processes operate on improved input data, leading to more reliable results in challenging conditions.
Data Source
Figure 1
Figure 2
Figure 3A~3D
AI summary
This invention provides an image processing apparatus which comprises a first correction unit which performs first gray-level correction on an image of a target region using a first correction amount, a first estimation unit which estimates the number of specific objects in the image of the target region corrected by the first correction unit, a second correction unit which performs second gray-level correction on the image of the target region using a second correction amount that is different from the first correction amount, a second estimation unit which estimates the number of specific objects in the image of the target region corrected by the second correction unit; and a selection unit which selects a larger one of the number of the specific objects estimated by the first estimation unit and the number of the specific objects estimated by the second estimation unit.