Image Processing Memory Segmentation for Concurrent I/O
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image processing apparatuses with small main memory sizes face inefficiencies in image data input and output due to the need for backup copies and limited resource utilization, as they cannot perform concurrent operations without significant performance degradation.
Innovation Solution
Implementing a dual input method based on image processing mode, where the sub memory serves as a paging or band buffer, and the main memory is used for temporary storage and backup copying of image data, allowing efficient page-by-page or band-by-band data handling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a main memory of large size is used to reserve memory area for image data, then quick image input and output can be performed concurrently, but the cost increases
Solution Approach 1:
The patent segments the memory system into three distinct components: sub memory for temporary storage, main memory for reserved areas, and non-volatile backup memory for permanent storage. This segmentation allows each memory component to serve a specific function, enabling concurrent operations without requiring the main memory to be excessively large. The sub memory handles immediate buffering while the main memory maintains reserved areas for both input and output operations.
Solution Approach 2:
The sub memory acts as an intermediary between the scanner and the main memory, and between the main memory and the non-volatile backup memory. It temporarily stores image data during transfer operations, allowing the main memory to reserve areas for concurrent operations without direct interference. This intermediary buffer enables efficient data flow management across the memory hierarchy.
2Quantity of substance
If a main memory of small size is used to reduce cost, then image data can be stored page by page, but image input on next page cannot start until backup copy is stored on backup memory device
Solution Approach 1:
The system performs preliminary actions by reserving memory areas in the main memory before image input operations begin. The main memory reserves areas for both current page processing and next page input, allowing the scanner to continuously read pages without waiting for backup operations to complete. This advance preparation eliminates bottlenecks in the image input process.
Solution Approach 2:
The patent implements a multi-level copying strategy: image data is copied from the scanner to the sub memory, then to the main memory, and finally a backup copy is made in the non-volatile memory device. This distributed copying approach allows the main memory to maintain small size while still enabling efficient operations through the sub memory buffer and reserved areas.
3Productivity
If a sub memory with paging buffer is used to enable quick image input, then image data can be obtained quickly from backup memory device, but device complexity increases
Solution Approach 1:
The main memory serves multiple functions simultaneously: it stores reserved areas for current page processing, maintains reserved areas for next page input, and coordinates with the sub memory and non-volatile backup memory. This multi-functionality reduces the need for separate dedicated memory components, simplifying the overall system architecture while maintaining high performance.
Data Source
AI summary
An image processing apparatus includes: an image input portion that inputs image data page by page; a sub memory; an encoder that encodes the image data; a main memory and a non-volatile memory; a mode judgment portion that judges a mode; and a controller that implements a first or second method depending on the judgment, the first method including: storing a part of one page on the sub memory; reserving a memory area for the one page in the main memory; storing the one page on the main memory; and reserving a memory area for next one page in the main memory, the second method including: reserving a memory area being capable of storing image data page by page in the sub memory; storing a part of image data on the main memory; storing a backup copy on the non-volatile memory; and deleting the image data from the main memory.