Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection
There are technologies available in research in order to have inspection on macro and micro cracks on eggshell. However, there are still some difficulties when coming to inspection on translucent areas where before the micro-cracks happening. Transfer leaning using pre-trianed neural network is used...
| Main Author: | |
|---|---|
| Format: | Monograph |
| Language: | English |
| Published: |
Universiti Sains Malaysia
2019
|
| Subjects: | |
| Online Access: | http://eprints.usm.my/58451/1/Pretrained%20Convolutional%20Neural%20Networks%20As%20Feature%20Extrator%20Of%20Eggshell%20Mottling%20Pattern%20For%20Quality%20Inspection.pdf |
| _version_ | 1797014072196071424 |
|---|---|
| author | Ng, Eng Yeong |
| author_facet | Ng, Eng Yeong |
| author_sort | Ng, Eng Yeong |
| collection | USM |
| description | There are technologies available in research in order to have inspection on macro and micro cracks on eggshell. However, there are still some difficulties when coming to inspection on translucent areas where before the micro-cracks happening. Transfer leaning using pre-trianed neural network is used at minimized computational resources while having a very high efficiency in classifying the eggs into three classes which are good, bad and unknown. Alexnet, Resnet and Inception of different architectures are compared to compute respective accuracy. It proved that the Alexnet gives highest predictive accuracy which is 96.80%, followed by Resnet, 93.15% and Inception, 90.16%. Results obtained from Alexnet is used to do statistical analysis such as ANOVA and student-t test to measure statistically significant differences between the means of accuracy from training set and testing set of image data. Visualization on channel along with activation strengths allow to know how a network learn to classify an egg with the help of Pareto chart. The deep dream images are generated by referring to the generation of images that produce desired activations. |
| first_indexed | 2024-03-06T16:10:59Z |
| format | Monograph |
| id | usm.eprints-58451 |
| institution | Universiti Sains Malaysia |
| language | English |
| last_indexed | 2024-03-06T16:10:59Z |
| publishDate | 2019 |
| publisher | Universiti Sains Malaysia |
| record_format | dspace |
| spelling | usm.eprints-584512023-05-11T04:17:11Z http://eprints.usm.my/58451/ Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection Ng, Eng Yeong T Technology TJ Mechanical engineering and machinery There are technologies available in research in order to have inspection on macro and micro cracks on eggshell. However, there are still some difficulties when coming to inspection on translucent areas where before the micro-cracks happening. Transfer leaning using pre-trianed neural network is used at minimized computational resources while having a very high efficiency in classifying the eggs into three classes which are good, bad and unknown. Alexnet, Resnet and Inception of different architectures are compared to compute respective accuracy. It proved that the Alexnet gives highest predictive accuracy which is 96.80%, followed by Resnet, 93.15% and Inception, 90.16%. Results obtained from Alexnet is used to do statistical analysis such as ANOVA and student-t test to measure statistically significant differences between the means of accuracy from training set and testing set of image data. Visualization on channel along with activation strengths allow to know how a network learn to classify an egg with the help of Pareto chart. The deep dream images are generated by referring to the generation of images that produce desired activations. Universiti Sains Malaysia 2019-05-01 Monograph NonPeerReviewed application/pdf en http://eprints.usm.my/58451/1/Pretrained%20Convolutional%20Neural%20Networks%20As%20Feature%20Extrator%20Of%20Eggshell%20Mottling%20Pattern%20For%20Quality%20Inspection.pdf Ng, Eng Yeong (2019) Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection. Project Report. Universiti Sains Malaysia, Pusat Pengajian Kejuruteraan Mekanik. (Submitted) |
| spellingShingle | T Technology TJ Mechanical engineering and machinery Ng, Eng Yeong Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection |
| title | Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection |
| title_full | Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection |
| title_fullStr | Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection |
| title_full_unstemmed | Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection |
| title_short | Pretrained Convolutional Neural Networks As Feature Extrator Of Eggshell Mottling Pattern For Quality Inspection |
| title_sort | pretrained convolutional neural networks as feature extrator of eggshell mottling pattern for quality inspection |
| topic | T Technology TJ Mechanical engineering and machinery |
| url | http://eprints.usm.my/58451/1/Pretrained%20Convolutional%20Neural%20Networks%20As%20Feature%20Extrator%20Of%20Eggshell%20Mottling%20Pattern%20For%20Quality%20Inspection.pdf |
| work_keys_str_mv | AT ngengyeong pretrainedconvolutionalneuralnetworksasfeatureextratorofeggshellmottlingpatternforqualityinspection |