Xedge>> LightSpeed>> 返回
项目作者： MasterT99

项目描述 ：
Car Damage Detection System using AI
高级语言： HTML
项目地址: git://github.com/MasterT99/LightSpeed.git
创建时间： 2020-09-18T19:34:47Z
项目社区：https://github.com/MasterT99/LightSpeed
开源协议：GNU Lesser General Public License v2.1
下载


LightSpeed

Car Damage Detection System using AI

Project Details

Team ID 86513
Project Title Car Damage Detection System
College Name L. D. COLLEGE OF ENGINEERING, AHMEDABAD
Department BE - Computer Engineering
Year 2020
Keyword Artificial Intelligence, Machine Learning, Deep Learning, Image Processing, Damaged Car Detection, Insurance Claiming
Internal Guide Prof. Pragnesh Patel

Team Details

  • Mehta Bhavy
  • Shah Keval
  • Shah Het
  • Shah Nigam
  • Tamboli Rushabh

Project Demo Video

https://youtu.be/B4H_FjYRPqg

How to run?

  1. Clone this repository.
  2. Create conda/pip vitrual environment with requirements.txt file.
    • Required module versions are listed in that same file of you want to install manually. Earlier or later versions may not work.
  3. Unzip model.zip in the directory LightSpeed\CDD\static\model.
    • Your directory should look like this:
      1.   LightSpeed\CDD\static\model
      2.   │   car.h5
      3.   │   door.h5
      4.   │   front.h5
      5.   │   hood.h5
      6.   │   rear.h5
      7.   │   side.h5
      8.   │   window.h5
      9.   │   winshield.h5
      10.   │
      11.   └───bumper
      12.       │   saved_model.pb
      13.       │
      14.       ├───assets
      15.       └───variables
      16.           │   variables.data-00000-of-00002
      17.           │   variables.data-00001-of-00002
      18.           │   variables.index
      19.           │
      20.           └───variables_temp_ea0f68758ae940398f50e26c2f256b68
      • Delete model.zip.001-015 files after successfully extracting model files.

  4. Create empty folders as shown below

    1. LightSpeed\CDD\static\upload
    2.          ├───coldstorage
    3.          │   ├───front
    4.          │   ├───left
    5.          │   ├───rear
    6.          │   └───right
    7.          └───predict
    8.              ├───front
    9.              │   └───a
    10.              ├───left
    11.              │   └───a
    12.              ├───rear
    13.              │   └───a
    14.              └───right
    15.                  └───a
    • Make sure to create folders, otherwise program will throw errors.

  5. Now change directory to root folder LightSpeed and execute python manage.py runserver

  6. Open http://127.0.0.1:8000 in your browser.
  7. Login or signup and upload your own car images or upload from given test image set from LightSpeed\CDD\static\upload\test folder.