Multi Model Processing System of Automatic Wiper Arm via Artificial Intelligence Based Prediction Module



Vooi, Vern Han (2021) Multi Model Processing System of Automatic Wiper Arm via Artificial Intelligence Based Prediction Module. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

[img] Text
VOOI VERN HAN_Full text.pdf
Restricted to Registered users only

Download (14MB)


Visual recognition has always been one of the useful and recognized applications to perform sorting techniques and in any manufacturing industry. This sorting process usually conducted by an operators manually, and thus faces a few problem due to human limitations. Therefore, it is necessary to design and build an automated visual recognition system to increase the efficiency. The goal of this final year project is to design an automatic visual recognition system used for model recognition and monitoring. Nippon Wiper Blade Co. (NWB) will funded the project by sponsoring their wiper product that were commonly used on automobile. In this project, three different sizes and types of wiper blade was studied which is Design Wiper, Neo-stream Wiper and Wiper Blade type NR27. Image captured by camera and their length and surface area will be used to serve as features to feed into AI to distinguish the differences between the model. Convolutional Neural Network (CNN) will be used to train the AI system by using MATLAB. The specific function of wiper recognition is considered to study the design and installation of the imaging system. Then, an visual recognition method is built which is a change-point detection method that serve as a parameters for a non-stationary recognition process. For an industrial, controlling the false alarm rate have to be considered so that unnecessary time is not wasted on the sorting process. The model parameters are estimated and used to design a statistical test to show the performance of the system and encourage the uses of visual recognition system for sorting techniques to replace human labor. Finally, the impact of illumination degradation on the distinguish detection performance is studied in order to predict the multi model needs for the imaging system. In conclusion, an image classification model will be implemented and use to differentiate three various types of wiper blade model sponsored by Nippon Wiper Blade Co. (NWB).

Item Type: Final Year Project
Subjects: Technology > Mechanical engineering and machinery
Technology > Electrical engineering. Electronics engineering
Faculties: Faculty of Engineering and Technology > Bachelor of Mechatronics Engineering with Honours
Depositing User: Library Staff
Date Deposited: 12 Jul 2021 09:09
Last Modified: 12 Jul 2021 09:09