Patentometric review on automated plant phenotyping

Plant phenotyping involves the measurement of observable traits of plants. Plant traits like leaf area, leaf count, leaf surface temperature, chlorophyll content, plant growth rate, emergence time of leaves, and reproductive organs depend on the interaction of its genotype with the environment. Plant phenotyping serves to analyze biotic and abiotic stresses on plants, select crop varieties resilient to the surrounding environment, and improve crop yield. Recent advancements in imaging technologies help expedite the growth of automatic, non-invasive, and efficient plant phenotyping systems. These plant phenotyping systems involve using different imaging techniques like visible imaging, hyperspectral imaging, chlorophyll fluorescence imaging (CFIM), thermal imaging to record, monitor, and analyze plant phenotypes using images. In the last few years, researchers have been working on developing image processing, computer vision, machine learning, and deep learning approaches for the accurate and precise analysis of plant images. Therefore, this paper reviews and presents insights about the research reported through patents in the area of automatic plant phenotyping. This review report uses patent databases like Espacenet, Lens, and Google Patents to search, review and analyze patent documents. The paper presents a patentometric analysis of all 67 patent documents available till date focusing on automatic image-based plant phenotyping. The review provides a summary and analysis of outstanding patents in terms of qualitative and quantitative patent indices. This article provides a comprehensive global patent study to aid researchers and scientists develop more efficient plant phenotyping algorithms, devices, and systems.