THE TRENCH CHANNEL DETECTION WITH IMAGE PROCESSING FOR AUTONOMOUS BOAT WATER SPRAYER

Abstract

Trench Channel agriculture is one of the easiest ways to manage crops in terms of water resources. Watering the crops requires the management of human resources which waste in human hours and labor cost. To reduce the waste and cost in human labor, boat and pump are integrated to be an equipment for watering crops in the field. This equipment be able to drive water in large quantities and efficiently. However, it still need human to maneuver the boat in the trench channel. For the intelligent agriculture, the equipment should be able to control autonomously and possibly to plan and manage without human control.
This research introduces the system to solve the autonomous navigation problem by using camera to detect the trench channel via machine vision by using image processing to determine the trench channel. The concept is to separate the trench channel by using difference color of ground and water to get the left and right trench channel lines. By using the intersection between the trench channel lines with the number of horizontal lines, the channel guideline can be calculated, which describes the direction of the trench channel. By comparing with the reference center point, the degree of maneuver can be determined and use as the command that send to the control the boat to maneuver autonomously.