【毕业设计】缺陷检测-基于深度学习实现的高效轮胎磨损+缺陷检测算法实现python源码.zip

# Efficient Tire Wear and Defect Detection Algorithm Based on Deep Learning Paper : https://www.koreascience.or.kr/article/JAKO202125761197586.pdf \ Project report : https://scienceon.kisti.re.kr/commons/util/originalView.do <br/> ## environment - cuda 10.2 - python 3.6.12 - torch 1.8.1 - torchvision 0.9.1 - tensorflow 1.14.0 - keras 2.1.0 ``` pip install -r requirements.txt ``` <br/> ## detectDefect This project is tire defect detection model using [@matterport Mask-RCNN balloon.py](https://github.com/matterport/Mask_RCNN.git). \ and start by reading this [blog post about the balloon color splash sample](https://engineering.matterport.com/splash-of-color-instance-segmentation-with-mask-r-cnn-and-tensorflow-7c761e238b46). ### 1. Create dataset Generate data Annotation and save JSON file using VGG Image Annotator, [VIA](https://www.robots.ox.ac.uk/~vgg/software/via/). Each mask is set of polygon points. ### 2. Download a pretrained model You can download pre-trained COCO weights (mask_rcnn_coco.h5) from the [releases page](https://github.com/matterport/Mask_RCNN/releases). ### 3. Train ``` python train.py --command train --weights coco --dataset [path/to/dataset] --image [path/to/image] ``` ### 4. Visualize Visualize bounding box and defect masks ``` python train.py --command splash --weights [path/to/checkpoint] --dataset [path/to/dataset] --image [path/to/image] ```  <br/> ## detectWear ### 1. Create tread mask Prepare images and Gaussian tread attention masks. The masks are generated by Mask-RCNN. <br/>  ### 2. Train ``` python train.py ``` ### 3. Test ``` python test.py --resume [path/to/checkpoint] ``` ### 4. Class Activation Map ``` python cam.py --dataset_path [path/to/dataset] --dataset [path/to/dataset/folder] --model_path [path/to/checkpoint/folder] --model_name [path/to/checkpoint.pth] ```