Entropy-Based Object Edge Curve Detection in Multispectral Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for object edge detection in images, such as those using edge trackers, are ineffective for objects with blurred or non-fixedly outlined contours, particularly in images from air-to-surface multispectral cameras.
Innovation Solution
A method that converts images to false color using the HSV color space, assigns hue values, classifies pixels as object or background, calculates an entropy profile using a displaceable evaluation window, and differentiates extreme values to determine object edge curves, enabling differentiation between artificial and natural objects and distinguishing monochromatic from polychromatic objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If edge trackers are used to detect object edges in images, then detection speed is improved, but detection accuracy deteriorates for objects with blurred or non-fixed contours
Solution Approach 1:
The patent transforms the image from RGB color space to HSV color space, changing the parameter representation of colors. This allows separation of hue information (which remains stable under lighting changes) from brightness information (which varies with lighting). By operating on hue values rather than raw pixel intensities, the method achieves both fast processing and accurate detection of objects with blurred contours.
Solution Approach 2:
The patent segments the image processing into distinct steps: first converting to HSV color space, then thresholding hue values to create a binary mask separating object pixels from background pixels. This segmentation allows the entropy calculation to focus only on spatial distribution of classified pixels, improving both speed and accuracy.
2Loss of information
If true color images are used for object detection, then color information is preserved, but differentiation between objects with similar colors deteriorates due to color mix-up
Solution Approach 1:
The patent changes the color space parameters from RGB (red, green, blue) to HSV (hue, saturation, brightness). The hue parameter in HSV represents the dominant wavelength of color and is invariant to brightness changes, allowing differentiation of objects with similar apparent colors but different spectral characteristics. This transformation eliminates the color mix-up problem while preserving essential color information.
Solution Approach 2:
The HSV color space acts as an intermediary representation between the raw RGB image data and the final object detection result. By converting to HSV and using hue thresholding, the method creates an intermediate binary mask that separates objects from background based on color characteristics, enabling accurate detection even when objects have similar colors in the original image.
3Stability of the object's composition
If hue-based classification is used to identify objects, then invariance to brightness fluctuations is improved, but detection capability for polychromatic objects deteriorates
Solution Approach 1:
The patent segments the object detection process into multiple hue ranges. For polychromatic objects, multiple hue thresholding operations are performed with different hue value ranges, creating multiple binary masks. These masks are then combined (logically ORed) to detect all parts of the object regardless of which color component they reflect. This segmentation approach maintains brightness invariance while extending detection capability to polychromatic objects.
Data Source
AI summary
In a method for determining object edge curves in a recorded image of a multispectral camera, the image recorded is converted into a false color image, hue value from the HSV color space is assigned to each pixel in the false color image such that the hue value corresponds to a hue angle on a predetermined color circle, and the pixels are classified as object pixels and background pixels, such that those pixels with hue values that are within a predetermined value range are defined as the object pixels. An entropy profile is then determined by way of a displaceable evaluation window, such that the entropy of mixing S is calculated for each pixel from the object pixels and background pixels according to the equation:S=-k(nAlnnAnA+nB+nBlnnBnA+nB)where nA denotes the number of object pixels within the evaluation window, nB denotes the number of background pixels within the evaluation window and k is a proportionality factor.