Comparisons of Various Image Preprocessing Approaches for Enhancing Medicinal Leaves.

PUSHPA, B. R.; ANUSHA, R.; BRUNDAVI, N.; DIVYA, L.; ISHWARYA, M. P.

Plants are widely used in foodstuff, medicines and industry and it's also essential for environmental protection. However, it's a crucial task to recognizes plant species on earth. Plants are depreciated due to pollution, diseased or nutrient deficiency. The current way of identification and determination of the type of medicinal leaves is still being done manually and prone to human errors. Misclassification of plant species results in huge loss on the production of crops and economical value of market. Plant recognition requires huge amount of labor, knowledge about the plants and also requires more time interval. Image processing is the method of study and manipulation of digitalized images especially to enhance the quality of the images. During image acquisition there is a chance of corruption due to noise images are corrupted by the noise that are applied on medicinal leaves. Filtering and smoothing of images are the important task to scale back or remove the effect of noise. This paper represents analysis of various image preprocessing techniques that are applied on medicinal leaf images. These techniques are used to remove the noise based on their potential. Image preprocessing is a main step for noise removal and for enhancing the standard of an original images. The quality of images is analyzed in support with PSNR (peek signal to noise ratio) values that are applied on various filtering and smoothening methods to get the better-quality images. Further these preprocessing methods are very useful for the segmentation process.

SIGNAL-to-noise ratio; PRODUCTION losses; IMAGE analysis; HUMAN error; PLANT species

International Journal of Pharmaceutical Research (09752366), 2020, Vol 12, Issue 3, p763

0975-2366

Academic Journal

10.31838/ijpr/2020.12.03.108