Abstract:
Now a days in farming Humanoid scarecrows are usually dressed and placed in open fields to discourage birds and animals from disturbing and feeding on recently cast seed and growing crops. Farmers used to make sounds, stay day and night in the fields it was an herculean task to them to stay in fields for long time. Scarecrow are not smart enough in night to discourage the birds and animals due to dark. Their is a need to advance and make them smart in this digital IoT era. Scarecrow plays a major role in fields to help in relaxing the farmers.
My present project smart scarecrow help the farmer to scare the animals and birds by activating sounding, blinking RGB colors, moving around the field when it identify an animal like wild boars, monkeys, dears, goats, cattle, etc and birds like pigeons, parrots, sparrows, crows, etc where are going to cause more damage during yields. The innovative idea was not implemented in anywhere so my project will be new compared to other devices which are existing in market.
Overall my project with Nordic Thingy:53 can help farmers and they like to place it in their agricultural land to protect crop in smart manner. It help in scaring birds or animals in apartments, harvesting centers, etc places also.
Edge impulse Project:
https://studio.edgeimpulse.com/public/148545/latest/
ML implementing:
Selecting the suitable feature to extract the classification.
Training and testing data for Cow, Rat, Dog, Cat, Monkey, Background (gb), Unknown.
Model testing result to identify the object.
Test Results:
The below images are training and testing images downloaded after completing the Machine learning. the interencing shows best results in classification and identification.
Below images are real-time identifying the animals on streets with close and far distance as the ML is running perfectly in Edge Impulse.
Deploying the ML on the Mobile is shown in the video.
Comments