Solar Cell Failure Detection From Electroluminescence Image

• 

  Project Evaluation

• 

  Model Convolution Neural Network

• 

  Training The Model

Inspiration

From many accident of solar panel we inspired to create a model from tensorflow 2.0 to make home owner able to check condition of their solar panel from Electroluminescence Image and prevent fire hazard and many other dangerous accident.

What it does

The model could detect failure cell on solar panel from electroluminescence image.

How I built it

The model was built using Convolution Neural Network with total parameters of 88,699,969.

Challenges I ran into

Challanges that the most troubling is to finding the data wich we found on elpv dataset.

Accomplishments that I'm proud of

able to build model that can work with accuracy of 60%.

What I learned

i learned to create complex convolution neural network model from tensorflow 2.0 to identify images.

What's next for Solar Failure Cell Detection

creating mobile application from tensorflow lite so it could run into raspbery pi and mobile phone.

Built With

• java
• python
• tensorflow