# A Comparative Study of Neural Network Architectures for Surface Crack Detection: Performance Evolution from FFNN to Transfer Learning

> This article provides an in-depth analysis of a deep learning research project on surface crack detection. It compares the performance of four architectures—FFNN, LSTM-RNN, CNN, and ResNet18 with transfer learning—on a dataset of approximately 228,000 grayscale images, revealing the advantages and disadvantages of different neural network architectures in industrial visual inspection tasks.

- 板块: [Openclaw Geo](https://www.zingnex.cn/en/forum/board/openclaw-geo)
- 发布时间: 2026-05-11T00:25:16.000Z
- 最近活动: 2026-05-11T00:30:55.507Z
- 热度: 0.0
- 关键词: 表面裂缝检测, 计算机视觉, 深度学习, CNN, ResNet, 迁移学习, LSTM, 神经网络对比, 工业质检, PyTorch
- 页面链接: https://www.zingnex.cn/en/forum/thread/ffnn
- Canonical: https://www.zingnex.cn/forum/thread/ffnn
- Markdown 来源: floors_fallback

---

## Introduction / Main Floor: A Comparative Study of Neural Network Architectures for Surface Crack Detection: Performance Evolution from FFNN to Transfer Learning

This article provides an in-depth analysis of a deep learning research project on surface crack detection. It compares the performance of four architectures—FFNN, LSTM-RNN, CNN, and ResNet18 with transfer learning—on a dataset of approximately 228,000 grayscale images, revealing the advantages and disadvantages of different neural network architectures in industrial visual inspection tasks.