Fast and Accurate Deep Learning Architecture on Vehicle Type Recognition

Vehicle Type Recognition has a significant problem that happens when people need to search for vehicle data from a video surveillance system at a time when a license plate does not appear in the image. This paper proposes to solve this problem with a deep learning technique called Convolutional Neural Network (CNN), which is one of the latest advanced machine learning techniques. In the experiments, researchers collected two datasets of Vehicle Type Image Data (VTID I & II), which contained 1,310 and 4,356 images, respectively. The first experiment was performed with 5 CNN architectures (MobileNets, VGG16, VGG19, Inception V3, and Inception V4), and the second experiment with another 5 CNNs (MobileNetV2, ResNet50, Inception ResNet V2, Darknet-19, and Darknet-53) including several data augmentation methods. The results showed that MobileNets, when combine with the brightness augmented method, significantly outperformed other CNN architectures, producing the highest accuracy rate at 95.46%. It was also the fastest model when compared to other CNN networks.

Electrical Properties of Semiconducting Copper Zinc Sulphide Thin Films

The electrical properties of metal-organic-chemical-vapour-deposited copper zinc sulphide (Cu-Zn-S) thin films on soda-lime substrates were studied. The films produced were characterized in terms of their electrical properties employing the Four-point probe procedure at a temperature range of 370 to 470°C. The electrical properties (resistivity and conductivity) of the deposited copper zinc sulphide films were systematically studied in terms of the deposition parameters of concentration and deposition temperature. The conductivity was in the interval of 5.48 to 8.0 × 10-1 (Ω.cm)-1. Activation energies of 0.54 and 0.29 eV in the deposition temperature range were estimated. The high resistive property of the films re-emphasized the potential use of these materials as active semiconductors for optoelectronic device applications.

Influence of Gasification Operating Parameters on Performance of the Nong Bua Dual Fluidized Bed Gasification System in Thailand

Gasification system performance generally depends on feed moisture content, activity of bed material, gasifier and combustor temperatures, and scrubber media. The tar concentration and gas composition of product gas are two indicators of the gasification system performance. In this research, the effects of gasifier temperature and the activity of bed material on the tar concentration and gas composition of the product gas produced from a dual fluidized bed (DFB) gasification system power plant were investigated. The DFB gasification system power plant is located in Nong Bua district, Nakhon Sawan province, Thailand. Two periods of gasification operation were examined. These two periods were when the olivine was freshy activated and then after a period of operation. The gasifier temperature had several peaks during the operation, which caused the product gas composition to fluctuate. When the olivine had been used for a period, the percentage of hydrogen was approximately 3% higher than when the olivine had been freshly activated, and a lower heating value was observed, which was probably due to lower heating value of hydrogen. The tar concentration was substantially lower when compared with the freshly activated olivine. When the olivine was used for a period, the average tar concentration was 56±22 mg/Nm3 (this is after 95 h continuous operating time) while the average tar concentration of the freshly activate olivine was 872±125 mg/Nm3 (which was after 34.5 h continuous operating time). It was concluded that the average tar concentration and gas composition were influenced by the activity of the bed material and the gasification temperature.