We live in a world where everything can be controlled and operated automatically, agricultural sector is not an exception. Due to the increasing demand in the agricultural industry, plants monitoring is seen as one of the most important task in any farming or agriculture based environment. The need to effectively grow plants and increase them is very critical. In order to do so, it is important to monitor the plants during their growth period. With the evolve of digital image processing techniques, there have been a rise in employing them to play an important role in many practical fields in our life. In this paper, we discuss an important and challenging task in image processing field, which is object tracking, to target the plants monitoring challenges. Object tracking can be defined as the process of extracting some useful sparse set of features to be tracked. As the method suggests, we implemented a set of techniques for extracting plants features to access important gardening knowledge.
The proposed techniques discussed in this paper are done in main four steps. The first step involves extraction of plants region using color spaces. The second step involves image enhancement that contains image preparation to be ready for feature extraction process. The third step involves feature extraction, where some features are extracted from the pre-processed image to help on the final step, by applying three methodologies, which they are: color intensity, structural differences measure and k-means cluster and transformed them into nominal values, finally, we applied analysis and interpretation for values to determine the growth rate, flowering and age stage cluster.
The test results from 52 sample of plants tracking show that the proposed methodologies are within the accepted level of accuracy, easy to understand and quick to solve, and practical to use. This research can be the base stone for researchers and students in the botany field to track different types of plants.