A Space Infrared Dim Target Recognition Algorithm Based on Improved DS Theory and Multi-Dimensional Feature Decision Level Fusion Ensemble Classifier

Xin Chen,Hao Zhang,Shenghao Zhang,Jiapeng Feng,Hui Xia,Peng Rao,Jianliang Ai

DOI: https://doi.org/10.3390/rs16030510

IF: 5

2024-01-30

Remote Sensing

Abstract:Space infrared dim target recognition is an important applications of space situational awareness (SSA). Due to the weak observability and lack of geometric texture of the target, it may be unreliable to rely only on grayscale features for recognition. In this paper, an intelligent information decision-level fusion method for target recognition which takes full advantage of the ensemble classifier and Dempster–Shafer (DS) theory is proposed. To deal with the problem that DS produces counterintuitive results when evidence conflicts, a contraction–expansion function is introduced to modify the body of evidence to mitigate conflicts between pieces of evidence. In this method, preprocessing and feature extraction are first performed on the multi-frame dual-band infrared images to obtain the features of the target, which include long-wave radiant intensity, medium–long-wave radiant intensity, temperature, emissivity–area product, micromotion period, and velocity. Then, the radiation intensities are fed to the random convolutional kernel transform (ROCKET) architecture for recognition. For the micromotion period feature, a support vector machine (SVM) classifier is used, and the remaining categories of the features are input into the long short-term memory network (LSTM) for recognition, respectively. The posterior probabilities corresponding to each category, which are the result outputs of each classifier, are constructed using the basic probability assignment (BPA) function of the DS. Finally, the discrimination of the space target category is implemented according to improved DS fusion rules and decision rules. Continuous multi-frame infrared images of six flight scenes are used to evaluate the effectiveness of the proposed method. The experimental results indicate that the recognition accuracy of the proposed method in this paper can reach 93% under the strong noise level (signal-to-noise ratio is 5). Its performance outperforms single-feature recognition and other benchmark algorithms based on DS theory, which demonstrates that the proposed method can effectively enhance the recognition accuracy of space infrared dim targets.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary

What problem does this paper attempt to address?

The paper attempts to address the problem of how to improve the recognition accuracy of spaceborne infrared dim targets in a strong noise environment. Due to the long distance and small size of space targets, and the lack of geometric texture information, relying solely on grayscale features for recognition may lead to unreliable results. Additionally, the imaging signal of the target is weak, and the feature information is easily affected by noise and other interferences, making it difficult to achieve high-precision recognition with a single feature.

To solve the above problems, the paper proposes a spaceborne infrared dim target recognition algorithm based on an improved Dempster-Shafer (DS) theory and a multi-dimensional feature decision-level fusion integrated classifier. Specifically, the method is implemented through the following steps:

1. **Feature Extraction**: Preprocess and extract features from multi-frame dual-band infrared images. The extracted target features include long-wave radiation intensity, mid-long-wave radiation intensity, temperature, emissivity-area product, micro-motion period, and speed.

2. **Feature Classification**: Input the radiation intensity features into the Random Convolutional Kernel Transform (ROCKET) architecture for recognition; use a Support Vector Machine (SVM) classifier for micro-motion period features; input other categories of features into Long Short-Term Memory (LSTM) networks for recognition.

3. **Evidence Fusion**: Construct the posterior probability of each classifier output using the Basic Probability Assignment (BPA) function of DS theory, and achieve target category discrimination through improved DS fusion rules and decision rules.

4. **Conflict Handling**: To address the counterintuitive results produced by DS theory in the presence of evidence conflict, introduce a contraction-expansion function to modify the evidence body to alleviate conflicts between evidence.

Experimental verification shows that the recognition accuracy of this method can reach 93% in a strong noise environment, which is superior to single-feature recognition and other benchmark algorithms based on DS theory, indicating that this method can effectively improve the recognition accuracy of spaceborne infrared dim targets.