Prioritized Image Data Layer Transfer for Visualization Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in efficiently visualizing large 3D digitalized image data with varying resolutions, leading to high latency and computing power requirements, especially when dealing with large datasets like those from microscopy or hyper-spectrograms, as they struggle to provide real-time visualization without significant resource consumption.
Innovation Solution
The system prioritizes the transfer of image data from an image data source to a sink by designating a reference layer with multiple resolutions, transmitting these in priority, and reloading other layers only when the reference layer reaches a preset resolution, thereby reducing data quantity and improving visualization efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all image data from multiple layers and resolutions are transmitted, then complete visualization is achieved, but data transmission volume and latency increase significantly
Solution Approach 1:
The patent segments image data into multiple layers (focus layers) with different depth information, allowing selective transmission of only necessary layers based on the region of interest, rather than transmitting all layers simultaneously
Solution Approach 2:
The system performs preliminary processing to identify the reference layer and determine which layers are actually needed for visualization, preparing a reduced dataset before transmission to minimize latency
2Adaptability or versatility
If image data is stored in multiple resolution levels, then zoom functionality is enabled, but data quantity increases significantly
Solution Approach 1:
Different layers store image data at different resolutions according to their specific needs - the reference layer stores multiple resolutions for zooming, while other layers store only the necessary resolution for their depth information, optimizing storage efficiency
3Ease of manufacture
If receiver-side or server-side conversion of large image data is performed, then data processing is achieved, but latency time and computing power requirements increase
Solution Approach 1:
The patent extracts and transmits only the essential image data from the reference layer at multiple resolutions, excluding redundant data from other layers, thereby reducing the computational burden on the receiving side while maintaining visualization quality
Data Source
AI summary
What is provided is a system for visualizing digitalized image data having an image data source and an image data sink. The image data are associated to at least two different layers, wherein, for at least one of the layers as the reference layer, image data is present in at least two different resolutions. The layers show a common scene at different conditions of taking a picture. The image data source is configured to transfer to the image data sink image data of the reference layer in the at least two different resolutions in a prioritized manner relative to the image data of layers different from the reference layer. The image data source is configured to transfer to the image data sink, in a reload case where the image data source has transferred the image data of the reference layer to the image data sink at a presettable reference resolution, the image data of a presettable number of layers different from the reference layer in at least one presettable reload resolution.


