微型电脑应用

光刻机在半导体制造中起着关键作用，但其性能易受多种因素影响，导致跟踪控制的精度和效率下降。为了提高光刻机Mark点检测的精度和稳定性，提升整体的光刻质量和生产效率，提出一种融合特征整合模型的改进基于边缘检测的更丰富卷积特征(RCF)算法。所提出的算法优化网络结构并重新设计损失函数，引入调制因子以增强算法对复杂样本的处理能力。实验结果表明，所提出的算法在600次训练时Mark点检测精度达90%。处理信号特征时，定位精度的真实值与预测值差距最大达到3.5 nm,显示出高准确性和稳定性。应用所提出的算法的光刻机材料完整度达97%,有效提升了光刻质量和生产效率。

The lithography machine plays a crucial role in semiconductor manufacturing, but its performance is susceptible to various factors, resulting in a decline in the accuracy and efficiency of tracking control. To improve the accuracy and stability of the Mark point detection in the lithography machine and enhance the overall lithography quality and production efficiency, an improved richer convolutional features for edge detection RCF algorithm integrating the feature integration model is proposed. The proposed algorithm optimizes the network structure and redesigns the loss function, introducing modulation factors to enhance the algorithm's ability to handle complex samples. Experimental results show that after 600 training iterations, the Mark point detection accuracy of the proposed algoritm reaches 90%. When processing signal features, the maximum gap between the true value and the predicted value of positioning accuracy reached 3.5 nm, demonstrating high accuracy and stability. The material integrity of the lithography machine using the proposed algorithm reaches 97%, effectively improving the lithography quality and production efficiency.