Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/486815Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Mohd Marzuki Mustafa, Prof. Dr. Ir. | - |
| dc.contributor.author | Azman Hamzah (P31646) | - |
| dc.date.accessioned | 2023-10-11T02:25:44Z | - |
| dc.date.available | 2023-10-11T02:25:44Z | - |
| dc.date.issued | 2012-11-29 | - |
| dc.identifier.other | ukmvital:81535 | - |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/486815 | - |
| dc.description | Kawalan kualiti adalah prosidur yang sangat penting dalam operasi pemprosesan makanan. Masa kini ia masih dilaksanakan secara manual oleh operator terlatih. Kebiasaannya, keputusan dibuat secara kualitatif iaitu bergantung kepada pengalaman, kecekapan dan ketekunan operator. Justeru, keputusan kualitatif yang dibuat secara manual ini kebiasaannya memberikan keputusan yang kurang tepat dan juga tidak konsisten bagi kelompok pemprosesan yang berlainan. Proses kawalan kualiti secara manual ini boleh diganti dengan penggunaan teknologi penglihatan komputer. Bagi pemprosesan dodol berenzim khususnya, tahap kritikal dalam kawalan kualiti ialah pengesanan detik mula penggelatinan buburan tepung pulut. Pengesanan detik mula penggelatinan yang tepat berupaya menghasilkan dodol berenzim berkualiti tinggi. Peranti kawalan seperti pemasa automatik atau pengawal suhu tidak berupaya menentukan detik mula penggelatinan buburan tepung pulut secara tepat semasa pemprosesan kerana faktor ketaksekataan kadar kehilangan haba kepada sekitaran melalui dinding luar alat pemasak dan juga permukaan alat pemasak yang terbuka. Kecerunan suhu diantara produk dan sekitaran yang tidak konsisten juga menyebabkan masa detik mula penggelatinan yang berbeza-beza. Oleh yang demikian kajian ini bertujuan memberi fokus kepada pembangunan algoritma pengesanan detik mula penggelatinan buburan tepung pulut menggunakan teknologi penglihatan komputer untuk mengautomasi proses pembuatan dodol berenzim. Pengesanan detik mula penggelatinan buburan tepung pulut menggunakan pendekatan penglihatan komputer telah dicadangkan supaya penentuan detik mula penggelatinan boleh ditentukan dengan tepat dan konsisten. Ia melibatkan pembangunan algoritma pemprosesan imej untuk mengesan detik mula penggelatinan buburan tepung pulut dan seterusnya mengautomasikan sistem pemasakan dodol berenzim secara masa nyata. Untuk menentukan teknik pengesanan detik mula penggelatinan terbaik, ujian telah dilaksanakan terhadap 30 kelompok pemprosesan buburan tepung pulut menggunakan enam algoritma pemprosesan imej iaitu (i) histogram, (ii) penolakan piksel imej, (iii) matriks keterjadian bersama aras kelabu (GLCM), (iv) kecerunan medan vektor (KMV), (v) kombinasi KMV dengan GLCM dan (vi) jelmaan Fourier pantas (FFT). Analisis dan prapemprosesan imej dilaksanakan terhadap perubahan tekstur imej buburan tepung pulut pada setiap 5 saat kerangka imej penggelatinan buburan tepung pulut semasa pemprosesan. Apabila algoritma pemprosesan imej dapat mengesan detik mula penggelatinan, maka bekalan elektrik kepada elemen pemanas diberhentikan serta-merta, untuk menghentikan proses pemanasan. Dapatan penyelidikan ini menunjukkan keupayaan algoritma pemprosesan imej terdapat sedikit perbezaan diantaranya tetapi teknik pemprosesan imej secara FFT memberikan keputusan terbaik dan amat meyakinkan iaitu memberikan purata keberkesanan pengesanan penggelatinan buburan tepung pulut 94.64% diikuti oleh teknik KMV 89.94%, penolakan piksel 87.27%, GLCM 86.19%, histogram 84.91% dan teknik kombinasi KMV dengan GLCM 69.78%. Kesimpulannya, teknik pemprosesan imej FFT didapati paling sesuai untuk tujuan khusus iaitu menentukan detik mula penggelatinan buburan tepung pulut. Ia juga telah memenuhi matlamat am kajian iaitu memantau dan mengawal kualiti semasa pemprosesan makanan secara masa nyata bagi menghasilkan kualiti makanan yang sentiasa konsisten walaupun di proses pada kelompok dan masa yang berlainan.,Quality control is a crucial procedure in food processing line. Currently the task is conducted manually by a trained operator. Typically, the decision is made qualitatively depending on the operator's experience, expertise and concentration. As such, these manually made qualitative decisions are normally imprecise and inconsistent for the different processing batches. The manual quality control process can be replaced by using computer vision technology. For the enzymatic 'dodol' processing in particular, detecting the gelatinisation onset of glutinous rice flour slurry is a critical step in quality control. The precise detection of the gelatinisation onset will help produce high quality enzymatic 'dodol'. Control device such as an automatic timer or temperature controller is unable to correctly determine the glutinous rice flour slurry gelatinisation onset during processing due to non-uniform heat loss to the surrounding through the cooker's wall and its opening. The surrounding and product temperature gradient also varies and these will result in variation of the glutinous rice flour slurry gelatinisation onset. Therefore this research is aimed to focus on the algorithms development for detecting the gelatinisation onset of glutinous rice flour slurry using computer vision technology for automating the enzymatic 'dodol' process making. The computer vision approach is proposed to detect the gelatinisation onset of glutinous rice flour slurry precisely and consistently. This involves the development of image processing algorithm for detecting the gelatinisation onset and also automating the cooking system for the enzymatic 'dodol' processing in real time. To determine the best detection technique for the gelatinisation detection, several testing were conducted on 30 batches of the glutinous rice flour slurry using six different image processing algorithms namely, (i) histogram, (ii) image pixels subtraction, (iii) grey level co-occurrence matrix (GLCM), (iv) vector field gradient (VFG), (v) combination of VFG with GLCM and (iv) fast Fourier transform (FFT). The image analysis and pre-processing was carried out on the image texture changes in every 5 seconds of the glutinous rice flour slurry image frame during processing. When the image processing algorithm detects the gelatinisation onset, the electric power supply to the heating element is switch-off immediately to stop the heating process. The research finding shows the varying results but the FFT technique gave the best and most convincing results which are the average of the effectiveness detection of gelatinisation early stage is 94.64%, followed by GFV 89.94%, image pixels subtraction 87.27%, GLCM 86.19%, histogram 84.91% and the combination of the GFV with GLCM 69.78%. In conclusion, the image processing technique, i.e., FFT was found to be the best technique for the specific aim of determining the gelatinisation onset of the glutinous rice flour slurry. It also fulfils the research general goals, that is, to monitor and control the quality of the food being processed in real time so as to consistently produce good quality food even though it was processed at a different batch and time.,Ph.D. | - |
| dc.language.iso | may | - |
| dc.publisher | UKM, Bangi | - |
| dc.relation | Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina | - |
| dc.rights | UKM | - |
| dc.subject | Penggelatinan | - |
| dc.subject | Pemprosesan makanan | - |
| dc.subject | Pengecaman imej | - |
| dc.subject | Food industry and trade -- Technological innovations | - |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | - |
| dc.subject | Dissertations, Academic -- Malaysia | - |
| dc.title | Penentuan detik mula optimum penggelatinan secara automatik berasaskan teknik pengecaman imej | - |
| dc.type | Theses | - |
| dc.format.pages | 269 | - |
| dc.identifier.callno | TP370.A995 2012 3 tesis | - |
| dc.identifier.barcode | 001647; 005651(2021)(PL2) | - |
| Appears in Collections: | Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_81535+SOURCE1+SOURCE1.0.PDF Restricted Access | 6.98 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.