Data Compression Apparatus for Object Detection Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data compression techniques fail to properly compress data for object detection processes, as they do not consider the characteristics of the object detection process, leading to inefficient compression of image and output data from monitoring apparatuses.
Innovation Solution
A data compression apparatus and method that sets a compression ratio based on the relationship between the compression ratio and detection accuracy, using a detection probability inference model generated by compressing data at various ratios and detecting objects, to optimize data compression for object detection processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is compressed using conventional techniques, then data size is reduced, but detection accuracy of object detection process deteriorates
Solution Approach 1:
The system performs preliminary actions by generating a detection probability inference model in advance through model generation apparatus. This model predicts detection accuracy for various compression ratios, allowing the data compression apparatus to select appropriate compression ratios before actual compression, thereby avoiding loss of detection accuracy while achieving data size reduction.
Solution Approach 2:
The system implements feedback by using detection probability inference models that predict detection accuracy based on compression ratios. The object detection process results feed back into the model generation, which continuously optimizes the compression strategy. This closed-loop feedback ensures that compression decisions are made with knowledge of their impact on detection accuracy, resolving the contradiction between compression and accuracy.
2Productivity
If compression ratio is increased to reduce data size, then transmission efficiency is improved, but detection probability of objects decreases
Solution Approach 1:
The model generation apparatus performs preliminary actions by pre-calculating detection probabilities for various compression ratios and storing these in detection probability inference models. This allows the data compression apparatus to make informed decisions about compression ratios that maintain detection probability while maximizing transmission efficiency, avoiding the need to test compression empirically.
Solution Approach 2:
The system changes parameters by dynamically adjusting compression ratios based on predicted detection probabilities. Instead of using fixed compression ratios, the system selects compression ratios that optimize the balance between transmission efficiency and detection probability for different data types and monitoring scenarios, resolving the contradiction through adaptive parameter selection.
3Speed
If conventional compression is applied without considering object detection characteristics, then compression speed is maintained, but compression effectiveness for detection purposes is insufficient
Solution Approach 1:
The system performs preliminary actions by pre-generating detection probability inference models that encode knowledge about optimal compression for object detection. This preliminary model generation allows the data compression apparatus to quickly select appropriate compression ratios without performing time-consuming empirical tests, thereby maintaining compression speed while improving compression effectiveness for detection purposes.
Data Source
AI summary
A data compression apparatus includes: a compression unit that performs a compression process on output data that is sequentially outputted as time series data by a monitoring apparatus for monitoring a monitoring target range; an output unit that outputs the output data on which the compression process is performed, to a data processing apparatus that performs an object detection process for detecting an object that exists in the monitoring target range by using the output data; and a setting unit that sets a compression ratio used in the compression unit, on the basis of an accuracy information indicating a relationship between a compression ratio at which the output data is compressed and a detection accuracy of the object by the object detection process performed by using the output data compressed at the compression ratio.