Date of Publication : June 2026

Abstract: Monitoring and maintaining ecological balance through monitoring plant growth is crucial for sustainable management of green spaces in urban areas. To achieve this, we build a Morphology-Based Integrated Plant Growth Monitoring System that uses digital image processing and machine learning to evaluate plant health from RGB photographs. The system includes four main components 1) Preprocessing, 2) Contrast Enhancement (CLAHE), 3) Plant Segmentation in HSV Color Space, and 4) Morphological Operations (such as Erosion, Dilation, Opening & Closing) to enhance the quality of the plant mask. A Vegetation Index (e.g. ExG), Texture, and Patch Area will be used to extract features from the images, and machine learning classification will be performed using a Random Forest Algorithm. The output from these different steps results in three classification categories for plant health Vegetation Health, Moderately Stressed Vegetation, and Vegetation that is Dry/Degraded. Testing this system on the park image data set yielded improved segmentation quality compared to those images processed without morphological processing – reduced amount of noise present in the image and increased connectivity of regions. The proposed model resulted in a classification accuracy of vegetative health assessment of about 80 % and also provides detailed maps of vegetation health and statistical report of the total area of each vegetation health category in the park. The proposed system will be computationally efficient and able to operate in real-time therefore, it is well suited to monitor and evaluate plant growth with standard devices.

Reference :