高级检索
    张毓, 高雅月, 常峰源, 谢将剑, 张军国. 小样本条件下基于数据扩充和ResNeSt的雪豹识别[J]. 北京林业大学学报, 2021, 43(10): 89-99. DOI: 10.12171/j.1000-1522.20210185
    引用本文: 张毓, 高雅月, 常峰源, 谢将剑, 张军国. 小样本条件下基于数据扩充和ResNeSt的雪豹识别[J]. 北京林业大学学报, 2021, 43(10): 89-99. DOI: 10.12171/j.1000-1522.20210185
    Zhang Yu, Gao Yayue, Chang Fengyuan, Xie Jiangjian, Zhang Junguo. Panthera unica recognition based on data expansion and ResNeSt with few samples[J]. Journal of Beijing Forestry University, 2021, 43(10): 89-99. DOI: 10.12171/j.1000-1522.20210185
    Citation: Zhang Yu, Gao Yayue, Chang Fengyuan, Xie Jiangjian, Zhang Junguo. Panthera unica recognition based on data expansion and ResNeSt with few samples[J]. Journal of Beijing Forestry University, 2021, 43(10): 89-99. DOI: 10.12171/j.1000-1522.20210185

    小样本条件下基于数据扩充和ResNeSt的雪豹识别

    Panthera unica recognition based on data expansion and ResNeSt with few samples

    • 摘要:
        目的  红外触发相机采集的雪豹监测图像质量参差不齐,且数量有限,为了提升小样本下雪豹的识别准确率,本研究提出一种雪豹监测图像自动识别方法。
        方法  该方法基于具备注意力机制的ResNeSt50模型,使用祁连山国家公园的雪豹监测图像作为原始数据集,红外触发相机拍摄的非雪豹陆生野生动物图像作为扩充负样本,网络雪豹图像作为扩充正样本,生成3种数据集并依次进行对比实验,选择合适的扩充方式引导模型逐步关注到雪豹个体关键特征,使用梯度类激活热力图可视化进一步验证数据扩充后的有效性。
        结果  使用原始数据集+扩充负样本+扩充正样本训练的模型识别效果最好,热力图可视化显示模型正确关注到雪豹个体花纹与斑点特征,对比基于Vgg16和ResNet50的识别模型,ResNeSt50的识别效果最好,测试集识别准确率达到97.70%,精确率97.26%,召回率97.59%。
        结论  采用本研究提出的原始数据集+扩充负样本+扩充正样本数据扩充方法训练的模型,可以区分背景与前景,且对雪豹本身特征具有较强的判别能力,泛化能力最好。

       

      Abstract:
        Objective  The quality of snow leopard monitoring images collected by infrared trigger cameras is uneven and the number is limited. An automatic recognition method of snow leopard monitoring images based on deep learning data expansion was proposed to improve the recognition accuracy of the snow leopard under limited samples.
        Method  Improving the ResNeSt50 model with attention mechanism, the snow leopard monitoring images of Qilian Mountain National Park of northwestern China were used as the original dataset, the non-snow leopard terrestrial wildlife images taken by the infrared trigger camera were used as the extended negative sample, and the network snow leopard images were used as the extended positive sample. Comparative experiments were conducted in turn based on the above three datasets. The model was gradually guided to focus on the key characteristics of individual snow leopards by choosing an appropriate expansion method, and the effectiveness of the data expansion was verified by Gradient-weighted Class Activation Map.
        Result  The model trained with the original data set+expanded negative samples+expanded positive samples had the best recognition effect. The Grad-CAM showed that the model correctly focused on the individual pattern and spot characteristics of the snow leopard. Compared with the recognition model based on Vgg16 and ResNet50, ResNeSt50 achieved the best recognition effect, the test set recognition accuracy rate reached 97.70%, the precision rate reached 97.26%, and the recall rate reached 97.59%.
        Conclusion  The model trained by the original data set+extended negative sample+extended positive sample data expansion method proposed in this paper can distinguish the background from the foreground, and has a strong ability to discriminate the characteristics of snow leopard itself, and the generalization ability is the best.

       

    /

    返回文章
    返回