Study On The Evaluation Of Detection Capability For SAR Image

The SAR image interpretation based on this technology is closely related to image quality.Due to its special imaging method,SAR image has many different characters from optical image.This makes it difficult to quantitatively measure the SAR image quality level from the delivered products.The existed image quality assessment method couldn't be effectively used for assessing the performance of image target detection,and is not helpful for the selection of SAR images and detection algorithms,which limits the application of SAR target detection.By utilizing the detection precision index to measure the capability and by analyzing the relationship between image detection results and basic image quality indicators,this paper studies the theories and techniques of image quality assessment for image target detection.Through the establishment of prediction model for image detection capability,the assessment of detection performance is achieved.The main work includes: SAR image quality evaluation,target detection algorithms for aircraft under single scene,and prediction model for image detection capability.This paper begins with the analysis of basic SAR image quality factors and detection algorithms.Based on the concept of "application suitability",the method of calculating detection capability index is provided to comprehensively evaluate the degree to which SAR image is suitable to certain target detection task.Theoretical and experimental results show that the constructed model provides both new ideas and effective tools for data capacity assessment in target detection applications for SAR image.Finally,a software platform for SAR image quality evaluation and detection capability testing is established.The software is capable of effectively evaluating SAR image quality,including basic objective indicator calculation,detection capability estimation as well as aircraft detection verification.