VDP Error Diffusion Initialization via Pixel Offset Padding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Error diffusion algorithms in high-speed applications face computational bottlenecks due to their sequential nature, particularly in the 'fill' and 'flush' processes required for vectorized-data parallel (VDP) error diffusion methods, which are computationally taxing and time-consuming.
Innovation Solution
The method initializes error buffers without performing the 'fill' and 'flush' processes by using a padding operation, where pixels behind the identified offset pixel in each scanline are set to a white value, allowing the multiple scanline error diffusion method to start immediately, thus eliminating the need for these operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the traditional fill and flush processes are performed for VDP error diffusion, then the error buffers are properly initialized and boundary artifacts are eliminated, but the computational cost increases and processing speed decreases
Solution Approach 1:
The patent extracts and eliminates the computationally expensive fill and flush processes from the VDP error diffusion method. By identifying that these processes are not strictly necessary when processing complete images or when boundaries are handled differently, the invention removes them to reduce computational overhead and improve processing speed while maintaining acceptable output quality.
Solution Approach 2:
The patent applies preliminary action by pre-initializing error buffers to zero before processing scanlines. This simple initialization replaces the complex fill and flush processes, establishing a known starting state for the error diffusion algorithm without requiring the computationally intensive boundary handling operations.
2Productivity
If multiple scanlines are processed in parallel using VDP error diffusion, then processing throughput increases, but the complexity of managing scanline boundaries and error diffusion artifacts increases
Solution Approach 1:
The patent segments the image processing into independent scanline batches that can be processed in parallel. By dividing the image into multiple scanline groups and processing each group independently with proper error diffusion within groups, the invention enables parallel processing while reducing the complexity of inter-scanline boundary management.
Solution Approach 2:
The patent applies local quality by treating different regions of the image with different processing approaches. Within each parallel scanline batch, full error diffusion is applied, while at the boundaries between parallel batches, simplified handling is used. This allows parallel processing to proceed with acceptable quality without requiring complex global boundary management.
3Measurement precision
If the error diffusion algorithm processes images sequentially line by line, then computational accuracy is maintained, but processing time increases for high-speed applications
Solution Approach 1:
The patent applies periodic action by processing images in periodic batches of scanlines rather than strictly sequentially. Multiple scanline batches are processed in parallel periodically, with error diffusion applied within each batch. This periodic batch processing maintains computational accuracy within batches while significantly reducing total processing time through parallel execution of multiple batches.
Data Source
AI summary
A method, non-transitory computer readable medium and apparatus for starting a multiple scanline error diffusion method are disclosed. For example, the method includes identifying a pixel for each scanline of a plurality of scanlines, wherein the pixel that is identified in the each scanline of the plurality of scanlines is offset, setting all pixels behind the pixel for the each scanline of the plurality of scanlines that is identified with a white pixel value and starting the multiple scanline error diffusion method.