XLXuchen Li

Xuchen Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Xuchen Li has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Identifying strawberry appearance quality by vision transformers and support vector machine

    by Hao Zheng, Guohui Wang, Xuchen Li

    The strawberry is a highly nutritional and beneficial cash crop, and its appearance quality sorting is a crucial step in the production process. Manual identifying and sorting, however, are subjective and more time consuming. In the field of image classification, vision transformers (ViT) have shown better performance. In this article, an extensive evaluation of the ViT models for the identification of strawberry appearance quality is presented. Moreover, to balance the accuracy and computation time of the ViT model, we propose a ViT‐based method that uses fine‐tuned ViT‐B/32 to extract the class token and imports it into the support vector machine (SVM) to identify strawberry. Besides, the class token is imported into the SVM with different kernel functions to evaluate their performance. The experimental results show that the highest recognition accuracy of original ViT models is ViT‐L/16, which can reach 97.38%. The application of the linear kernel function is more suitable in this work. The accuracy of original ViT‐B/32 is 93.7% and the proposed method improves the accuracy by 4.4% up to 98.1%. Furthermore, the required computation time of the proposed method is only 62.13 s, which is faster than other models. Therefore, the proposed method demonstrates enhanced robustness and universality, especially for abnormal and ripe strawberries.Practical Applications: The identification of strawberry appearance quality is a labor‐intensive task. Traditional methods mainly rely on manpower, which has high cost and low efficiency. Therefore, this work uses the fine‐tuned ViT‐B/32 to extract features and import them into SVM to identify the appearance quality of strawberries, which obtains the identification results more accurate and efficient. This study could facilitate the development of smart picking equipment for strawberries.


  • Recognizing strawberry appearance quality using different combinations of deep feature and classifiers

    by Xuchen Li, Changlong Cai, Hao Zheng, Hongfei Zhu

    The quality of the strawberry appearance is the most important indicator for consumers and modern fruit process engineering, which is closely associated with the ripeness and diseases. The continuous development of profound deep learning has greatly helped the recognition of strawberry appearance quality, granting a vigorous tool of relative precise outcomes, but the better performance of deep learning requires more time and more computation for training. This paper analysis the efficiency of different combination models that utilize deep feature plus classifiers for recognizing strawberry. The six convolutional neural networks are used to extract the deep feature and then the feature is imported into eight classifiers for classification. The average accuracy of six combination models is 1.53% higher than transfer learning. Besides, a method combining ResNet101 plus linear discriminant analysis (LDA) is proposed. The results show that the accuracy of the proposed method is 96.55%, which is superior to the transfer learning. Moreover, the training time of the method is 57.70 s faster than transfer learning. Therefore, this method has positive significance for the development of recognizing strawberry appearance quality.Practical Applications: The goal of our study is to make it easier for farmers to accurately classify the appearance of strawberries in real situations. This paper analysis the efficiency of different combination models that utilize deep feature plus classifiers for recognizing strawberry. Meanwhile, we have proposed a method combining ResNet101 with linear discriminant analysis (LDA), which obtain the classification task more accurate and efficient. This method has an important role in recognizing strawberry, which will facilitate the development of efficient sorting equipment and provide a viable strategy for future smart sorting methods. Therefore, this work has practical application value.


  • A FPGA-based method for modeling and generation of the time-varying electromagnetic environment

    by Xuchen Li, Donglin Su, Li Li, Zhipeng Xu, Shaohua Yang

    The frequency hopping radio has been widely used for its anti-interference ability. However, in the installation, testing and evaluation of short-wave frequency hopping radio, a realistic time-varying electromagnetic environment support is lacking. Hence, it is important to construct the time-varying electromagnetic environment which the short-wave frequency-hopping radio may face. To solve this problem, this paper firstly finds out live kinds of signals commonly faced by the short-wave frequency-hopping radio and extracts their time-varying parameters by signal simulations with MATLAB. After that, a model based on the traversal is proposed, which accurately reconstructs the time-varying electromagnetic environment. Finally, this model is implemented on the FPGA, which could generate a feasible time-varying signal to represent the electromagnetic environment. This device may be useful for the installation, debugging, testing, evaluation and other tasks of short-wave frequency hopping radio.