Authors:
Suganthi M, J. G. R. Sathiaseelan

Addresses:
1,2, Department of Computer Science, Bishop Heber College, Affiliated to Bharathidasan University, Trichy, Tamil Nadu, India. sugan.swt09@gmail.com1, jgrsathiaseelan@gmail.com2 

Abstract:

In this paper, image mining techniques are used for the purity test of various X-Ray seeds. Analyses of physical purity provide information on the percentage of pure intrinsic structure in a seed lot. An image mining technique has established a software application that can forecast seed pictures for seed lots. People may readily snap digital images anywhere, anytime, using a camera or mobile phone equipment, thanks to advances in camera technology. Additionally, using a computer system makes the transformation and processing simple. This research examines various image-mining methods that minimize the time and effort necessary to evaluate seedling growth performance while improving measurement accuracy. On the acquired X-ray image of the seed image, pre-processing techniques such as de-noising and feature extraction is performed to detect purity. Several denoise algorithms have been introduced, each with its benefits and controls. When choosing the right denoising algorithm, we need a good understanding of seed morphology. This paper presents a comparative analysis of four filter techniques. Feature extraction is associated with the seed's spatial, color, shape, texture, and statistical features. In order to classify various seeds using a feature extraction technique to produce the best results, this research developed a new texture feature extraction method.

Keywords: Guided Filter; Gaussian Filter; Median Filter; Feature Extraction; Reform Conventional Filter (RCF); Peak Signal to Noise Ratio (PSNR); Mean Square Error (MSE); Structural Similarity Index Metric (SSIM).

Received on: 05/11/2022, Revised on: 11/01/2023, Accepted on: 26/02/2023, Published on: 23/03/2023

FMDB Transactions on Sustainable Health Science Letters, 2023 Vol. 1 No. 1, Pages: 41-53

• Views : 845
• Downloads : 11
• Cited by: 22

Download PDF