Compression Method For Optical Characteristic Data Analysis

With the rapid development of sensor,huge optical data acquisition for various apps comes true,which provides essential data protection for object detection,identification and analysis.While,the huge optical data makes its transmission more difficult and also has storage problem.Therefore,an effective way of batch data compression and fast decoding is needed.In light of this issue,by analyzing optical data features,this paper proposes an efficient method of data compression based on sparse representation and run-length coding.This paper proceeds from the following four aspects:1.Analyze optical data and extract the main part by studying its features and rearrange the data for subsequent compression.2.study respective characteristics of each optical data by visual analysis.so as to provide a basis for selection and design of compression methods.3.Corresponding samples are needed when using sparse representation to compress.When there is a large amount of data,random sampling is adopted to save training time.When the sample size is small,use GAN network to generate training samples.4.Propose a hybrid coding method based on sparse representation and lossy run-length coding,use after training dictionary to segment and compress data and write the data into compressed file in a certain form.Different optical data have different characteristics.This paper designs different compression methods for corresponding optical data based on the characteristics of them.the optical data compression method based on sparse representation well balances the recovery effect and time efficiency,which helps the optical data storage and its subsequent application analysis.