lit waste not wilds

lit waste not wilds
Project Code in Jupyter Notebook
Project Code in Visual Studio
Website
Github Commits
Github different branches and merging them with main branch
Future Enhancement

Inspiration💡

In recent times, Fire🔥🔥 accidents have been significant disasters. Fire🔥🔥 accidents cause major damage to human life. To decrease the damage caused due to fire🔥🔥 accidents we came up with the Image processing Technology. This project can be implemented in any crowded place. In the current scenario, fire🔥🔥 outbreak is a common issue happening these days all over the world and the damage caused by this type of incident is tremendous toward nature and causes loss of human life. As per the research report given by Accidental Deaths and Suicides in India (ADSI), it was stated as fire-related accidents have killed 35 people every day on average in the last six years between 2016 and 2021. To overcome this issue, we have come up with a safe and cost-effective model. Though there are existing technologies for fire🔥🔥 detection, our proposed system can be deployed easily in the existing cameras like CCTV which makes it portable and reliable. As there are no hardware components involved in this model, hence it is cost-effective and will be much more efficient in nature.

What it does🚩

In this, we are using image processing techniques which is compatible with surveillance devices like CCTV, and wireless camera to UAVs. It can be mostly used in Railway stations, crowded places like markets, and bus stations, since the cameras are already installed in such places, and this system is aimed at diminishing the disadvantages of false alarms which makes the system cost-effective, efficient, and a fast method for detecting fire🔥🔥. Coming to the working model, whenever the fire🔥🔥 is detected it sends a message to nearby fire🔥🔥 stations and to the concerned departments with a beep or alarm sound at the place where the fire🔥🔥 outbreak has occurred.

How we built it🧱⚙️

Libraries Used📚📚

• NumPy • OpenCV • SMTP • Play sound • Threading

Challenges we ran into🧗✨

The first challenge was to import the libraries or the packages of Python and use an IDE to get the output since we were getting module not found errors and pip not recognized as internal/external command errors in cmd. We tried to execute our code on Visual Studio Code, Anaconda(Jupyter Notebook). We even tried to execute it on the command prompt of Windows. The second challenge was to choose a method out of all possible methods. The third challenge was to use Gaussian Blur we got many errors while using it. The last challenge was with HSV since it was difficult to get those lower and upper boundaries. The next challenge was to set a count(25000) through which if we get a count value above it alarm sound is triggered. The problem here was for any orange or yellow color objects the alarm was getting triggered(Ex:- An orange sketch) so we had to do more research and then decide the count value for our project.

Accomplishments that we're proud of🏆✨

• Can expect accurate output • Improving technology • Autonomous

What we learned🏫✨

We learned about OpenCV, HSV, and GaussianBlur. We learnt about different OpenCV methods such as resize(), waitkey(), etc. We even learned how to install different libraries of python and use it by importing those into our code.

What's next for lit waste not wilds🔮

We can develop this project in a more modern way by using different paths. We can re-develop this project by using temperature sensors and capturing the exact position of fire🔥🔥 and in many ways. There is much more way to develop this and make this more useful in the future.

Alternate Methods🗝️🗝️

1)By using Y Cb Cr: YCbCr (also Y'CbCr, Y Pb/Cb Pr/Cr) is a family of color spaces used in videos and digital photography that separates brightness (luma) from chroma (color). Green is naturally included as part of brightness because it's the color people are most sensitive to.

2)Cyan Magenta Yellow Key(CMYK): RGB is best for websites and digital communications, while CMYK is better for print materials. Most design fields recognize RGB as the primary colors, while CMYK is a subtractive model of color.

3)Convolutional Neural Networks: A convolutional neural network (CNN) is a type of artificial neural network used in image recognition and processing that is specifically designed to process pixel data.

4)Thermal Detection System: The infrared camera-based FireTIR system captures the temperature distribution in an area and automatically detects the hot spots. The thermal imaging cameras of the FireTIR system are calibrated and they get in real time the temperature information in each pixel, automatically detecting any outbreak of fire🔥🔥.

5)CIE L*a*b* Color Space