Image Processing Time Prediction Using Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for predicting the period required for image processing, such as the RIP process, face inaccuracies due to variations in conditions between processing different parts of image data, making it difficult to accurately estimate the processing time for entire image data based on partial data.
Innovation Solution
An information processing device and method that acquire attribute information and setting information related to image data, using a machine-learned model to predict the image processing time by incorporating data on the performance of the image processing device, allowing for more accurate predictions without needing to understand the specific correlations between data attributes and processing times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If image processing time is estimated from partial data processing, then prediction speed is improved, but prediction accuracy deteriorates
Solution Approach 1:
The system performs preliminary analysis by extracting multiple attributes (image size, resolution, color space, compression ratio) and processing conditions (RIP settings, rendering mode, dithering method) before actual processing. This preliminary characterization enables accurate prediction without executing the full processing operation, thus maintaining both speed and accuracy.
Solution Approach 2:
The patent changes from estimating based on single partial processing to analyzing multiple parameters including image attributes (size, resolution, color space), processing conditions (RIP settings, rendering mode), and device performance characteristics. This multi-parameter approach significantly improves prediction accuracy while maintaining computational efficiency through the established model.
2Device complexity
If simple partial data processing is used for estimation, then device complexity is reduced, but prediction accuracy deteriorates
Solution Approach 1:
The system segments the image processing prediction problem into distinct components: image attribute extraction (size, resolution, color space), processing condition analysis (RIP settings, rendering mode), device performance measurement, and model-based synthesis. Each segment handles specific aspects independently, maintaining system simplicity while achieving comprehensive accuracy through the integrated model.
Solution Approach 2:
The patent introduces an intermediary prediction model that acts as a mediator between raw processing parameters and final time estimation. This model synthesizes information from multiple sources (image attributes, processing conditions, device performance) without requiring complex direct measurement of actual processing time, thus reducing system complexity while maintaining accuracy.
Data Source
AI summary
An information processing device includes an information acquisition unit that acquires at least one of attribute information of image data to be image processed and setting information related to the image process on the image data, and a prediction unit that predicts, based on a machine-learned model that accepts input of at least one of the attribute information and the setting information to output a period required for the image process, a period required for the image process on the image data from information acquired by the information acquisition unit.