Image Processing Apparatus Edge Preservation via Digitizing Epsilon Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques using low-pass filters tend to dull edge portions due to averaging pixel values, leading to increased processing load and inefficiency in preserving edges while blurring other areas.
Innovation Solution
An image processing apparatus that performs gradation conversion, generates level-value-limited conversion images using digitizing epsilon filters with different level value ranges, synthesizes these images, and performs inverse gradation conversion to produce a smoothed image with reduced processing load and preserved edge details.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a low-pass filter is used to smooth an image, then the smoothing effect is achieved, but edge portions become dim and gradients get dull
Solution Approach 1:
The patent applies different processing treatments to different regions of the image based on local characteristics. Edge portions are identified and processed differently from non-edge regions, allowing edges to preserve their sharpness while other areas receive smoothing treatment. This is achieved through edge detection mechanisms that distinguish between edge and non-edge pixels, enabling localized quality control.
Solution Approach 2:
The image processing is segmented into distinct operations: edge detection, edge portion identification, and selective smoothing. By dividing the processing into separate stages and applying different algorithms to different segments (edge vs. non-edge regions), the patent avoids the uniform smoothing that causes edge degradation in conventional low-pass filtering.
2Manufacturing precision
If conventional epsilon filter is used to preserve edges, then edge portions are accurately saved, but processing load increases significantly
Solution Approach 1:
The patent extracts only the essential edge information from the image using edge detection, then applies smoothing operations selectively. Instead of processing every pixel through complex conditional branches as in conventional epsilon filters, the method extracts edge portions and applies targeted processing, reducing the overall computational load while maintaining edge preservation.
Solution Approach 2:
The patent performs preliminary edge detection and identification before applying smoothing operations. By pre-identifying edge portions and marking them for special treatment, the subsequent smoothing process can proceed more efficiently without needing to evaluate complex conditions for each pixel, thereby reducing processing time and computational burden.
3Manufacturing precision
If low-pass filter averages pixel values, then smoothing is achieved, but processing efficiency decreases due to high computational load
Solution Approach 1:
The patent applies smoothing operations only to non-edge regions identified through edge detection, rather than uniformly processing the entire image. This localized approach maintains smoothing quality in areas where it is needed while avoiding unnecessary computations in edge regions, thereby improving processing efficiency without sacrificing smoothing effectiveness.
Data Source
AI summary
An image processing apparatus performs a gradation conversion on an input image (frame) input from the outside to generate a conversion image and performs a plurality of smoothing processes corresponding to a plurality of preset level values on the generated conversion image or a reduced image generated from the conversion image to generate a plurality of level-value-limited conversion images. Then, the image processing apparatus selects one or a plurality of conversion images from the plurality of level-value-limited conversion images, synthesizes the selected images, generates a synthesis image, performs the gradation conversion on the generated synthesis image, and generates a smoothed image that is converted to have the same gradation as that of the input image.