Week 9

Activity: Research on previous project and methodology

This week I continue doing literature review on fire detection. The paper that I read is “Machine Learning Method for Image Recognition-based Fire Detection System”. In this journal, the authors combine multiple sensor systems which is temperature sensor, smoke sensor and CO sensor with image recognition to develop fire detection system. In this research paper, they used Convolutional Neural Network (CNN) are used as fire image feature extraction. The aim of this research paper is for the authors to compare between K-nearest neighbors (KNN) and linear logistic classification for fire image classification. From this research, I concluded that both machine learning classification algorithms can obtain a relatively high correct detection probability if training photos are sufficient. 

After finishing literature review, I continue my project’s methodology. This week I had to choose the software, hardware and algorithms that I want to use for my project. First, I decided to use OpenCV for the library. OpenCV is a popular computer vision library that includes many machine learning algorithms for image and video processing.

For IDE, I chose Visual Studio Code. Visual Studio Code is a source code editor developed by Microsoft for Windows, Linux, and macOS. 

For face recognition, I chose Haar Cascade as face detection classifier and Local Binary Histogram Pattern (LBPH) as face recognition algorithm. For fire detection, Convolutional Neural Network as fire detection algorithm.