A Survey on Image Analysis Techniques in Agricultural Product

Background/Objectives: In this paper we have analyzed various image analysis techniques and their issues in seed technology. The crucial role of these techniques is identification and classification of seeds, grading of seeds, quality determination of seeds in seed science and food processing sectors. We provide a complete survey of image analysis techniques and propose a processing module for seed identification and classification. Methods/Statistical analysis: We identified the problem of manual identification and classification of seed in agricultural sector. We provide five important processing steps as a proposed methodology for identification and classification of seed. By using this technology, we able to discriminate the defected seed from normal seed. Our proposed methodology includes noise removal, edge detection, segmentation and classification for seed identification. These methods provide a significant way for seed identification. Findings: In this work, we implemented a five processing module for seed identification. Seed image has taken for the basis of image acquisition and then a seed image is pre-processed by noise removal and image enhancement. Enhanced image goes through the processes of edge detection and segmentation. From the segmented image we extract features like colour, shape and texture for normal and defected seed which may help to identify the seed by image analysis techniques. Application/Improvements: Our work is useful to know about various issues in seed identification in agricultural sector. This work provides efficient methods to overcome the manual identification of seeds. We can also extend this work with an efficient identification and classification methodology in image analysis techniques in agricultural sector