X-Ray Image Decomposition Model for Cargo-Vehicle Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image recognition systems for cargo inspection face challenges in accurately distinguishing cargo from vehicle structures due to overlapping information in X-ray images, leading to reduced recognition accuracy and increased risk of missed detections, and require large amounts of labeled data that are difficult to obtain.
Innovation Solution
A method involving an image decomposition model trained with a first and second adversarial neural network, utilizing generation and discrimination networks, to separate cargo and vehicle images through a training process that adjusts parameters based on loss function values determined by inputting and comparing training images, employing negative logarithmic processing and image fusion techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine intelligent recognition is used for cargo detection, then inspection efficiency is improved, but recognition accuracy deteriorates due to overlapping cargo and vehicle body information in X-ray images
Solution Approach 1:
The patent applies segmentation by dividing the complex X-ray image into separate components: cargo region and vehicle body region. The neural network model processes the overlapped image to segment and identify distinct regions, allowing accurate detection of cargo while separating it from vehicle structure information, thus resolving the accuracy-efficiency contradiction.
2Measurement precision
If supervised training with large amounts of labeled data is used, then model accuracy is improved, but data acquisition difficulty and time consumption increase
Solution Approach 1:
The patent employs data augmentation techniques that create synthetic copies and transformations of limited real X-ray images. By generating virtual samples through geometric transformations, noise addition, and composite image synthesis, the model achieves high accuracy without requiring extensive real labeled data, thus reducing data acquisition time while maintaining model performance.
Data Source
AI summary
A method of training an image decomposition model, a method of decomposing an image, an electronic device, and a storage medium are provided. The method of training the image decomposition model includes: acquiring a training set; inputting first and second training images into first and second adversarial neural networks respectively, so as to determine a first loss function value; inputting a third training image into the first and second adversarial neural networks respectively, so as to determine a second loss function value; determining a third loss function value according to a comparison result between an acquired fusion image and the third training image, where the fusion image is generated by fusing generated images of the first and second adversarial neural networks, and adjusting a parameter of the image decomposition model according to at least one of the first to third loss function values.