The aim of this study is to investigate the parameters that could be used to identify abnormal gait pattern in Parkinson's disease subjects during normal walking. Hence, three types of gait parameters namely basic, kinematic and kinetic are evaluated. Initial findings showed that the average mean of cadence, step length and walking speed for Parkinson's disease patients are lower than normal subjects, while the mean of stride time for Parkinson's disease patients are higher. Further, for kinematic parameter, overall joint angle of hip, knee and ankle mean values are lower for Parkinson's disease patients as compared to normal group. In addition, for kinetic parameter, all mean values of ground reaction force parameters are higher for normal subjects with walking speed contributed as the major determinant. To evaluate the significant features that could be used as identification between PD and normal subjects, statistical analysis is conducted. Hence, based on the statistical analysis results, it was found that step length, walking speed, knee angle as well as vertical parameter of ground reaction force are the four significant features as indicators for classification of subject with Parkinson's disease based on the accuracy attained with Artificial Neural Network as classifier.

Combined Support Vector Machine (SVM) and Principal Component Analysis (PCA) was used to recognize the infant cries with asphyxia. SVM classifier based on features selected by the PCA was trained to differentiate between pathological and healthy cries. The PCA was applied to reduce dimensionality of the vectors that serve as inputs to the SVM. The performance of the SVM utilizing linear and RBF kernel was examined. Experimental results showed that SVM with RBF kernel yields good performance. The classification accuracy in classifying infant cry with asphyxia using the SVM-PCA is 95.86%.

No abstract

An increasingly connected production in the sense of Industry 4.0 allows completely new possibilities in regard to improved and more efficient production and higher product quality. But a key factor to Industry 4.0 is a consistent data flow along the production chain. However, the exchange of data, especially between companies, still is a major obstacle to overcome in order to achieve the aforementioned advantages. Currently, there are increasing efforts to record and analyse data. But there is a lack of a holistic system to handle data, therefore commonly company databases or other inefficient methods are used. These solutions are limited with regard to data exchange since the ownership of data cannot be proven, production data has no unforgeable timestamp, which in turn hinders the generation of complete production history from the final product (e.g., car door) back to the semi-finished product (e.g., steel sheet). As a result, there is insufficient to no data exchange along the production chain. In order to solve these problems blockchain is a promising approach. At the Institute of Manufacturing Technology, an operational blockchain system was developed and implemented using standard production machines. With the combination of a quarto rolling mill and a 400t - press, representing the sheet metal supplier and a forming company, respectively, the typical process chain of sheet metal processing is represented, which allows the detailed investigation of the established blockchain in this field of application. Within this contribution, the conceptual approach of a blockchain system for forming technology will be presented. The nature and the classification of occurring data throughout the production chain will be addressed.

This paper discusses the potential of applying VGG16 model architecture for plant classification. Flower images are used instead of leaves as in other plant recognition model because the structure of shape and color of leaves are similar in nature. This might be disadvantageous when we want to use only leaves images as a sole feature of plants to classify the species. Previous work has demonstrated the effectiveness of using transfer learning, dropout and data augmentation as a method to reduce overfitting problem of convolutional neural network model when applied in limited amount of images data. We have successfully build and train the VGG16 model with 2800 flower images. The model able to achieve a classification accuracy score of 96.25% for training set, 93.93% for validation set and 89.96% for testing set.

No abstract

No abstract

This project presents a method to detect diabetic retinopathy on the fundus images by using deep learning neural network. Alexnet Convolution Neural Network (CNN) has been used in the project to ease the process of neural learning. The data set used were retrieved from MESSIDOR database and it contains 1200 pieces of fundus images. The images were filtered based on the project needed.  There were 580 pieces of images types .tif has been used after filtered and those pictures were divided into 2, which is Exudates images and Normal images. On the training and testing session, the 580 mixed of exudates and normal fundus images were divided into 2 sets which is training set and testing set. The result of the training and testing set were merged into a confusion matrix. The result for this project shows that the accuracy of the CNN for training and testing set was 99.3% and 88.3% respectively.