Theses & Dissertations

Permanent URI for this collection

https://hdl.handle.net/20.500.14170/209

Browse

Recent Submissions

Now showing 1 - 5 of 366
  • Publication
    DCT image compression implemented on raspberry Pi to compress image captured by CMOS image sensor
    (Universiti Malaysia Perlis (UniMAP), 2019)
    Ibrahim Saad Mohsin
    An image contains large amount of digital data and it is necessary to reduce digital data volume for transmission and preservation by using image compression. This project mainly concentrates on image compression using a Raspberry Pi processor, which helps to preserve a large number of images and in retaining its quality. Signal, Image and data compression are an important signal processing tool in many areas of research, such as medicine, biometric, telecommunication, automotive and. The purpose of compression is to reduce the amount of data at the same time maintain the quality of image and signal for the other purpose. DCT transform is a family of image compression where the raw image is transformed to the other domain to produce smaller size of data. DCT transform has low computational complexity and fast processing algorithm. In this project, DCT transform will be implemented using Raspberry Pi SBC development board running on an ARM based processor. The raspberry Pi board has an advantage of image processing implementation due to the existing software development tool offered a rich feature for image processing such as OPENCV. The project consists of several design stages such as image pre-processing, the development of DCT algorithm, error rate computation and measurement by PSNR and MSE. The algorithm is developed using Python programming language with the additional image processing library such as OpenCV and NumPy mathematical library. The development of Internet and multimedia technologies that grow exponentially, resulting in the amount of information managed by computer is necessary. This causes serious problems in storage and transmission image data on real time. Therefore, should be considered a way to compress data so that the storage capacity required will be smaller. In this research wanted to know the influence of DCT to the compression ratio and to the PSNR (Peak Signal to Noise Ratio). Then the compression ratio and PSNR results on real time provide excellent visual quality. The result of applying DCT compression algorithm on images with six compression rate level which are 10, 20, 50, 100,170 and 200. The best performance can be achieved with compression rate level 200. However, on reducing the quality level of compression rate, the error measurements start becoming worse until a point is reached, where the perceptual difference from the original image can be easily noted.
  • Publication
    VAPT: VANET adaptive power transmission nodes connectivity
    (Universiti Malaysia Perlis (UniMAP), 2019)
    Siti Asilah Yah
    Recently, vehicular Ad-Hoc Network (VANET) has received considerable attention where its safety applications and Internet users accessing related application are expected to be supported. Safety applications play an important role to avoid road accidents. Even in the case if the accident is unavoidable, these applications can at least minimize the impact of accidents. The main goal ofV ANET is to provide safety to passengers and the real-time safety message exchanging is one of the most important issues in VANETs. Safety applications are a delay-sensitive and they mainly rely on reliable inter-vehicle communication. In VANET, the rapid and frequent changes in vehicle mobility create a highly dynamic topology. Due to this, VANET faces challenges in maintaining the connectivity. Therefore, by using fixed transmission power in VANET, it will give adverse effects on the connectivity. Transmission power is directly proportional to transmission range. Using fixed transmission power/transmission range leads to many disadvantages; for example, when vehicles are distributed densely, more nodes have to share the medium, causing more contentions, collisions, resulted to delays that reduce network capacity. In addition, the contention window ( CW) and bit rate (BR) also play a significant role in the transmission process. Using a bigger size of CW will reduce the packet loss, hence increase the throughput and Packet Delivery Ratio (PDR). However, the bigger size of CW will increase the delay. While for BR, the delay can be minimized by using a higher BR, but it will reduce the PDR and throughput. This is due to the fact that higher BR resulted in fast transmission rate which increases the packet loss. Hence, this research work is proposed in order to overcome these problems. The main objective of this research is to propose a new power control mechanism to maintain network connectivity that can minimize the delay and the uses of energy and improve the PDR. The proposed VANET Adaptive Power Transmission (VAPT) mechanism modified the MAC layer by applying oflline optimization using Differential Evolution (DE) algorithm to inspect three parameters (transmission power, BR, and CW) and to fmd their optimum value. These optimum values are then used to optimize the network performances. To achieve the objective of this research, extensive experiments were conducted to investigate how the variable value of transmission power can influence the connectivity of the network in different VANET scenarios and traffic load parameters. The results show that the proposed VAPT optimization mechanism improves the performance of multiple VANET scenarios. The highway scenario in different packet size simulation shows the highest performance improvement in terms of delay and energy consumption. The delay is improved by 55.07% in average and the energy consumption is improved by 54.44% in average. While for the PDR, the highest improvement is 17.35% in average is shown by hybrid scenario in different packet rate simulations. Then, the highest improvement for throughput is 17.78% in average, also shown by the hybrid scenario in different packet rate simulations.
  • Publication
    Analysis on non-destructive mango ripeness classification system
    (Universiti Malaysia Perlis (UniMAP), 2019)
    Muhamad Farid Mavi
    Malaysia is a country that receives enough sunlight and rain throughout the year forstrong fruit production, thus this unique property enables many fruits especially mangos to be cultivated. Mango obliges to be harvested at the best time to produce high quality of fruits and thus ripeness assessment are the most important aspects. However, with the increasing demand by consumers, the current way of mango ripeness assessment or classification is still prone to human error and sometimes damage the fruit because it is being done through manual processes. Therefore, the mango ripeness classification technique is required for replacing these traditional methods since this will improve the efficiency and accuracy. This research introduced a non-destructive system for mango ripeness classification by combining three techniques in a single system using image processing technique, odor sensing and capacitive sensing technique. The processes in this system are comprised with webcam camera for capturing the mango images forripeness classification through color based on the hue, saturation and value features through image processing technique. Besides that, eight different features were taken from Odor Gas Sensor (OGS) that are used to extract the smell data release from the mango, whereas as for capacitive sensing, parallel capacitive plate is used as an enhancement to the system for measuring the capacitive value features. The multiple technique is capable to tackle the problems arise as some mango has different ripeness properties such as color, odor and dielectric. Support Vector Machine (SVM) is used as classifier for this system to classify the mango ripeness and voting-based scoring approach is used to evaluate the overall performance score based on the three techniques. Based on the results and analysis, the mango ripeness classification system produces good results with a classification accuracy of 98.43% through image, odor and capacitive combination. The conducted research suggests that this system which are different from the existing system has the potential to implement for other types of fruits and can contribute to the improvement in the agriculture sector.
  • Publication
    FPGA prototyping flow optimization by using failed path fixes and manual clock distribution techniques
    (Universiti Malaysia Perlis (UniMAP), 2019)
    Salahuddin Savugathali
    The purpose of this research is to review and propose a new flow in FPGA Prototyping Flow using a Synopsys Protocompiler tool to solve the stated problem. A fascinating property of a latch-based design is that the combinational path delay is allowed to be longer than the clock cycle as it can borrow time from the shorter paths in the subsequent logic states. Time Borrowing technique is a common approach used to satisfy timing violation in an FPGA prototyped design. However, based on previous studies, more efforts are required to produce an efficient result in closing the timing violation, where lead us to propose a Failed Path Fixes technique. This approach is meant to fix the failed path in a latch due to the gated clock conversion (GCC) process during the synthesis stage which could lead to the timing violation. A solution for prototyping a multi-million logic gates of ASIC/SoC circuit into the FPGA platform for verification purpose is by partition the design into multi-FPGA. This research is focusing on the required partition requirement to successfully prototype the large SoC circuit into the multi-FPGA. The presence of cut clocks in a circuit after partition stage will result in the failure in routing stage due to the congestion error. Therefore, an approach proposed in this research to resolve this challenge is Manual Clock Distribution technique, so that design is able to meet the partition requirement to complete the prototyping process into multi-FPGA. The combination of our Failed Path Fixes and time borrowing technique are able to solve the timing violation problem by eliminating the unnecessary path created by protocompiler tool. Comparison of numbers of negative slack before and after our proposed technique is applied resulting 90% improvement. The manual clock distribution technique proposed has been able to solve the cut clock issue that leads to routing congestion problem when partitioning a circuit into two FPGA chips. With our proposed technique, all partition requirements are met and I 00% cut clock elimination is achieved.
  • Publication
    Enhancement of host-based and network-based mobility management protocols in wireless mesh network
    (Universiti Malaysia Perlis (UniMAP), 2018)
    Hoh Wei Siang
    Since the dependency of human toward the Internet, this has rapidly raise the usage of internet which may cause congestion and intermittent connection issues. The Wireless Mesh Network (WMN) was introduced to overcome the intermittent connection issues and caters rural area in order to effectively increase the scalability of network with lower construction cost. However, in WMN, there is no specific mobility management protocol that is appointed to handle the location management and handoff management of Mobile Node (MN). Besides that, the MN also faces the difficulties in choosing the suitable mobility management protocols to perform the handover process either in inter or intra domains since the current mobility management protocol handles both intra and inter domains mobility management. Single mobility management protocol needs to handle beacon update and acknowledgement for inter and intra mobility managements. This causes significant delay and distortion which decrease the efficiency of wireless communication. Having known this issue, the aims of this thesis are: firstly, performance evaluation and comparison of various mobility management protocols in WMN environment. The Host-Based and Network-Based mobility management protocols are consolidated in WMN environment and separately implemented and handles the inter and intra mobilities. MIPv6 and FMIPv6 are chosen to operate in intra network scenario while for inter network scenario, HMIPv6, FHMIPv6 and PMIPv6 are appointed to handle the operation. The perfonnance results show that the FMIPv6 have better perform than MIPv6 in intra network scenario. While for inter network scenario, the PMIPv6 is outperformed which compares to HMIPv6 and FHMIPv6. Secondly, the target is to enhance the performance of standard FMIPv6 in intra network scenario. The Optimise FMIPv6 (O-FMIPv6) was introduced with new fast handover scheme based on the standard FMIPv6 to operate in intra network scenario. For enhanced fast handover mechanism, when the MN senses lower signal strength, MN advertises to the neighbour network for the need to attach to the new higher signal strength access point. MN informs the new access point of the need to change to the new access point before the process of handover. Thus, this can reduce the handover latency of handover processes. Thirdly, it is to propose route optimise PMIPv6 (RO-PMIPv6) with new route optimisation scheme which modifies theLMA and MAG entitles. When there is a need to connect to different hierarchical access points, only LMA involves to inform the different hierarchical access points and CN before and after the handover process is performed. This can reduce the time of handover process and thus decreases the handover latency and increases the wireless communication efficiency. For both intra and inter domains mobility management over WMN, the end-to-end delay, packet delivery ratio (PDR) and throughput are measured to show the optimisation of the proposed O-FMIPv6 and RO-PMIPv6. Lastly, the results show that the designed intra domain mobility with O-FMIPv6 and the inter domain mobility with RO-PMIPv6 enhanced the mobility management performances. The O-FM1Pv6 performs 2.2% lower in end-to-end delay, 3.7% higher in PDR and 5.4% higher in throughput as compared with standard FMIPv6. The RO-PMIPv6 performs 58.3% lower in end-to-end delay, 2% higher in PDR and 2. 7% higher in throughput as compared with standard PMIPv6. With this enhanced mobility management, it is believed that with the implementation of the designed mobility management protocols, this can reduce latency, increase throughput and decrease distortion. With these criteria, the future aims of wireless communication while moving is made possible which eases the communication between human.