Detection and Image Classification System is a platform for forestry officer to monitor trail camera from accommodation of the National Reserved Forest.
ProcessConvolutional Neural Network (CNN) is one of deep learning method. Sending picture and initial label then throw in the retraining script to get model. When we get new image from trail camera [by connecting via bluetooth and tranfer over TCP protocol], importing image to classification script. This script will tell proability of each label. If the urgent label get high proability, it will send over LoRa technology by using Near Real-time Image Transfer System for National Reserved Forest.
ExplanationThis platform divide to 2 part.
1.1) Image Recieve Chart : Real-time update for new incoming images.
1.2) Daily Image Amount Chart : Show amount of detect in each interval of the day.
1.3) Weekly Image Amount Chart : Show amount of image in weekly to see the trend of invade.
2.1) Image Card : Show image with prediction accuracy from model and time when recieve the image.
2.2) Click Image : Image will move to the gallery.
