Gas Features Cooperative Calculation Neural Network Combined With an Electronic Nose System for Peanut Origin Identification

The misrepresentation of low-quality peanut as high-quality varieties is a common issue in the market. To address this problem, this study proposes a Gas Features Cooperative Calculation Neural Network (GFCC-Net) combined with an electronic nose (e-nose) system for the quality identification of peanut from different origins. Using the e-nose system, gas information from peanut of various origins is collected and analyzed. Considering the data characteristics of the gas information, the Gas Features Cooperative Calculation module (GFCC) is proposed. This module integrates both local and global deep features computations, extracting local features via convolutional operation and global features using a self-attention mechanism. Based on the proposed GFCC, a lightweight GFCC-Net is designed. The network undergoes structural optimization, and its performance is validated through ablation studies and comparisons with state-of-the-art classification methods. Experimental results demonstrate that GFCC-Net achieves the superior classification performance, with an accuracy of 98.07%, a precision of 97.95%, and a recall of 98.29%. These findings highlight the effectiveness of the proposed GFCC-Net in combination with the e-nose system for peanut origin traceability, offering an effective technical solution to ensure peanut quality and uphold market standards.