A Prototype of Raspberry Pi Camera Morphometric Measurements on Sheep Using Digital Image Processing

Abstract

Quantitative data are needed to identify and predict opportunities for increasing livestock productivity. The increase in livestock productivity can be assessed from the dimensions of the body of the livestock by measuring directly using morphometric. Morphometric data is an important parameter used to study livestock anatomy, productivity, growth rate, and performance quality of livestock. This research is focused on how to make a prototype development of a camera tool that can measure livestock morphometrics at a distance. The Camera use in mini camera with a raspberry pi as a microcontroller. And method to process the capture of data is Artificial Neural Network that downloaded from tensor flow. The result of this experiment show that the accuracy of the system is 50%. However, the camera has not been able to identify the object of farm animals correctly. There are several livestock objects detected as dog or cat. For the future work the library must be modified so it can only detect the sheep or goat so the morphometrics can be done.