Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476167Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Siti Norul Huda Sheikh Abdullah, Professor Dr. | |
| dc.contributor.author | Layla Wantgli Shrif Amosh (P53607) | |
| dc.date.accessioned | 2023-10-06T09:14:16Z | - |
| dc.date.available | 2023-10-06T09:14:16Z | - |
| dc.date.issued | 2013-02-04 | |
| dc.identifier.other | ukmvital:74747 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476167 | - |
| dc.description | Oil palm trees grow in the tropics and originated in Africa. In 2008, Malaysia and Indonesia produced nearly 55% of world production and 62 % of world exports. Besides rubber and rice, oil palm or Elaeis Guineensis remains as one of the most important plantation crops in Malaysia. Unfortunately, the lack of experience in oil palm fruit grading among the plucking farmers results in wrong estimation when harvesting. This affects production, negatively. Meanwhile, region growing with conventional image segmentation techniques need manually or fixed initial seed selection which, actually, increases the computational cost, as well as, implementation time. Hence, two main goals of this study are to enhance seed region growing for segmentation oil palm fruit image and to develop oil palm fresh fruit bunches (OPFFB) plucking system. This study presents n-Seed Region Growing (n-SRG) for color image segmentation by choosing adaptive numbers of seed. The proposed n-seed region growing technique is based on pre-defined number of regions and its gray scale histogram as the initial and its subsequent seeds. The proposed framework for Oil Palm Fresh Fruit Bunch (OPFFB) plucking has six phases comprising the following: color image capturing, image transformation from color image to gray scale, image segmentation using proposed n-seed region growing algorithm, and finally image ripeness classification using ratio color index production rule. The data sample consists of 240 images which comprises two ripeness classes ('ripe' and 'unripe'), and 'easy' and 'difficult' categories collected from Universiti Putra Malaysia benchmark dataset and Kuantan, correspondingly. Our proposed work is compared with k-mean clustering subsequently method and achieved 86% and 80% on easy case while difficult cases were 81.42% and 73.5%. To conclude, the proposed work has out-performed k-mean clustering method about 83.71% and 76.75% of average accuracy rates, correspondingly.,Master/Sarjana | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
| dc.rights | UKM | |
| dc.subject | Oil palm | |
| dc.subject | Seed | |
| dc.subject | Image segmentation | |
| dc.title | Automated n-seed region growing for oil palm fresh fruit bunch plucking system | |
| dc.type | theses | |
| dc.format.pages | 81 | |
| dc.identifier.callno | TA1638.4.A486 2013 3 | |
| dc.identifier.barcode | 000344 | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_74747+Source01+Source010.PDF Restricted Access | 2.85 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.