VRU Research and Development Journal Science and Technology

EFFECTS OF GREEN PEA, PIGEON BEAN, AND FAVA BEAN PROTEIN ISOLATES ON THE PHYSICAL CHARACTERISTICS AND NUTRITIONAL VALUE OF JACKFRUIT SAUSAGE PRODUCTS

2025-03-19T16:11:31+07:00

This study was aimed to develop and evaluate the physical and chemical properties as well as nutritional value of plant-based sausage products made from young jackfruit, supplemented with protein isolates extracted from green peas (Green Pea Protein Isolate: GPPI), and pigeon peas (Pigeon Pea Protein Isolate: PPI), and fava beans (Fava Bean Protein Isolate: FBPI). The primary focus was on improving texture and nutritional value to be similar to meat sausages. The experimental results were indicated that young jackfruit exhibits high antioxidant potential, comparable to 193.56 milligram of ascorbic acid per 100 grams and contains a total phenolic content of 3.76 milligram per gram/gram, highlighting its nutritional benefits. The analysis of physical properties were revealed that plant-based sausages supplemented with pigeon pea protein exhibited the highest lightness (L*) value of 53.11 and had appropriate hardness and chewiness. The plant-based sausages supplemented with fava bean protein had the highest moisture content (p<0.05) at 69.02%, with a firm and elastic texture. Meanwhile, the plant-based sausages supplemented with green pea protein demonstrated moderate moisture and protein content but had a lighter and more easily chewable texture. Consumer acceptance testing was indicated that the plant-based sausages supplemented with green pea protein received the highest overall acceptability scores for taste, texture, and overall preference, with no significant difference from the control formulation (p≥0.05). In contrast, the plant-based sausages with fava bean protein supplementation were highly scored for aroma (p<0.05) but requiring improvements in appearance and color. The plant-based sausages supplemented with pigeon pea protein isolated received sensory acceptance scores at a slightly satisfactory level. Therefore, green pea protein isolate showed strong potential for the development of plant-based sausages that were made from young jackfruit, as it provided an optimal texture and high nutritional value. The current study has been highlighted the feasibility of utilizing plant-based proteins in the development of high-quality, consumer-accepted alternative food products.

COMPARING THE PERFORMANCE BETWEEN FP-GROWTH AND APRIORI ALGORITHMS FOR ANALYZING SHOPPING PATTERNS IN A COFFEE SHOP

2025-03-19T12:40:50+07:00

The current study, namely Comparison of FP-Growth and Apriori Algorithms for Analyzing Shopping Patterns in a Coffee Shop, was aimed to investigate association rules and compare the performance of the Apriori and FP-Growth algorithms in analyzing customer purchase transactions in a coffee shop. The dataset, consisting of 1,000 transactions, was obtained from the Kaggle website then analyzed using data mining techniques with the Python programming language and the mlxtend library. The research results were revealed that both algorithms could generate association rules with high confidence values. For example, the rule (Latte) → (Croissant) achieved a confidence of 85%. However, FP-Growth algorithms demonstrated the better performance than Apriori’s, spending 0.01 seconds of processing time and using 30.34 KB of memory. On the other hand, Apriori algorithms spending 0.00 seconds and using 30.33 KB. Although Apriori algorithms are suitable for small datasets, FP-Growth offers a structural advantage through its use of the FP-Tree, which reduces redundant data scans. FP-Growth algorithms are more suitable for large and complex datasets.

ARTHROPOD COMMUNITIES IN THREE DIFFERENT AGRICULTURAL PRODUCTION SYSTEMS IN WANG NAM KHIAO, NAKHON RATCHASIMA PROVINCE, THAILAND

2025-04-24T13:51:49+07:00

Agricultural management has a significant influence on a large proportion of arthropod species, negatively impacting the community structure of arthropods, as reported in modern intensive agriculture. Our research aims to explore differences in arthropod species diversity and abundance between agricultural areas and their surrounding environments under three different agricultural practices and to investigate the relationships between arthropod diversity and local impact factors such as climate and agricultural practices. Arthropod diversity and abundance were sampled using pitfall trapping in three types of agricultural areas and their adjacent zones. In total, 99 morphospecies were identified within the study area. Greater richness values were recorded for the organic farming system (OM) compared to good agricultural practice (GAP) and conventional agricultural sites (CH). The number of species was higher in the inside zones than in the outside zones across all study sites. Significantly higher species richness in the inside zones compared to outside zones was observed at the OH and GAP (P<0.05), whereas no significant difference was found at the CH (P>0.05). Additionally, in the OM and GAP areas, but not in the CH, these differences suggest that the field edges of agricultural practices can play an important role in maintaining biodiversity in agroecosystems, and this role is related to edge-of-field practices in agriculture.

STREAMFLOW SIMULATION BASED ON LAND USE/COVER CHANGE USING INTEGRATED GOOGLE EARTH ENGINE AND SWAT+ IN AN AGRICULTURE-DOMINATED BASIN, NORTHEAST THAILAND

2025-03-03T17:22:50+07:00

Streamflow is a fundamental component of the hydrological cycle, and land use/land cover (LULC) significantly influences runoff processes. This study investigates the impacts of Land Use and Land Cover Change (LUCC) on streamflow in the Upper Songkhram River Basin (USRB), Northeast Thailand using Google Earth Engine (GEE) and the SWAT+ hydrological model. GEE facilitated high accuracy LULC classification (overall accuracy: 83-91%, Kappa coefficient: 0.76-0.85), revealing a marked increase in para rubber plantations and built-up areas, and
a decrease in paddy fields and forests between 2003 and 2023. The calibrated and validated SWAT+ model (NSE: 0.86/0.79, R²: 0.91/0.89, PBIAS: -24.5%/-36.7%) simulated the streamflow changes associated with LUCC. Results indicate a substantial increase in wet season streamflow, particularly in August (28.9 m³/s) and September (25.6 m³/s), primarily due to the decline of paddy fields and their water retention capacity. This study emphasizes the link between land use and hydrology, showcasing the combined utility of GEE and SWAT+ for assessing LUCC impacts. These findings offer valuable insights for sustainable water resource management and land use planning in Thailand and comparable regions worldwide.

PERFORMANCE COMPARISON OF CONVOLUTIONAL NEURAL NETWORK-BASED DEEP LEARNING MODELS FOR RHIZOME IMAGE CLASSIFICATION IN THE ZINGIBERACEAE FAMILY

2025-04-16T11:53:34+07:00

Herbal plants have been an essential part of traditional medicine for centuries. They have been used to treat various ailments and promote overall health. Despite significant advances in modern medicine, herbal plants have played a crucial role and been extensively employed across numerous sectors. A major challenge in their application is the accurate identification and classification of species, as herbs within the same family often exhibit remarkably similar physical characteristics. Such similarities can cause misclassification, leading to ineffective products or potential health risks. This study evaluated and compared four convolutional neural network models including ResNet-50, VGG-19, DenseNet201, and InceptionV3 for classifying rhizome images of four species in the Zingiberaceae family: Turmeric, Zedoary, Plai, and Wild turmeric. This aimed to properly adjust their parameter and evaluated their efficiency of model. The researchers collected and publicly released a dataset of 2,111 images then applied data augmentation to increase training diversity. After that, hyperparameter tuning was performed to optimize model performance. Experimental results were demonstrated that DenseNet201 with 200 training epochs, outperformed the other models, achieving the highest classification accuracy. These findings were suggested that the proposed model has been highly suitable for practical use in herbal industries, aiding species identification, product standardization, and quality control.

DEVELOPMENT OF WOOD-PLASTIC COMPOSITE PANELS FROM MARIAN PLUM BRANCH RESIDUES AND POLYETHYLENE FOR COMMUNITY ENTERPRISES

2025-05-08T11:37:37+07:00

The primary objective of this research was to develop wood-plastic composite panels produced from plastic-biomass composites, utilizing Marian plum branch residues, a natural material, and polyethylene plastic, which is widely used in producing industrial materials. The project was aimed to fabricate material suitable for use in community enterprises that require multifunctional applications and practical materials. In the current study, the researchers designed ten different mixing ratios of Marian plum branch residues and polyethylene to determine the most suitable composition for manufacturing composite wood panels. The wood-plastic composite panels gained were tested for mechanical properties according to the Thai Industrial Standards TIS 2998-2562 and TIS 876-2547, which relate to wood-plastic composite panels produced from plastic-biomass composites.

Test results indicated that the optimal wood-plastic composite panels should be consisted of a ratio of 3.5 parts of polyethylene plastic to 1 part of Marian plum branch residues by weight. The panel achieved a density of 877.35 kilograms per cubic meter, a bending strength of 14.10 megapascals, a modulus of elasticity of 2,112.98 megapascals, and a perpendicular tensile strength of 0.95 megapascals. These properties were suggested that the developed synthetic wood panel has been well-suited for indoor applications, particularly for interior decoration, where only moderate mechanical loads are encountered.

CAUSAL FACTORS INFLUENCING QUALITY OF LIFE AMONG PREGNANT WOMEN SUPHAN BURI PROVINCE

2025-05-27T14:26:28+07:00

The quality of life of pregnant women is critical as it affects adequate prenatal care, which is a key factor in reducing maternal mortality rates. This research study employed a descriptive cross-sectional design using structural equation modeling (SEM) and path analysis to investigate the factors influencing the quality of life among pregnant women in Suphan Buri Province, Thailand. The research sample comprised 400 pregnant women selecting through purposive sampling from four hospitals, which were chosen by ranking the number of pregnant women from highest to lowest and using simple random sampling. The questionnaire was the research instrument, which was divided into six sections, including both open-ended and closed-ended questions. Data analysis involved descriptive statistics, confirmatory factor analysis (CFA), and path analysis using statistical software.

The research findings revealed that many pregnant women had an overall quality of life at a high level, accounting for 70.75%. Pregnancy health literacy, health promotion motivation, and social support had direct effects on health promoting behaviors with standardized path coefficients of 0.56, 0.21, and 0.11, respectively. Moreover, they had indirect effects on quality of life with coefficients of 0.48, 0.18, and 0.10, respectively. Health promoting behavior had direct effect on quality of life, with a coefficient of 0.86. These factors collectively explained 65% of the variance in quality of life (R² = 0.65). The developed structural equation model demonstrated good fit with empirical data, indicated by fit indices: Chi-Square = 415.71 df = 211 Chi-Square/df = 1.97 CFI = 0.97 GFI = 0.98 AGFI = 0.96 RMSEA = 0.04 and SRMR = 0.04.

The research findings highlighted that pregnancy health literacy, motivation for health promotion, social support, and health-promoting behaviors have been key determinants influencing the quality of life among pregnant women. Therefore, public health agencies should develop programs focusing on strengthening these components to enhance the quality of life among this population.

CARBON FOOTPRINT ASSESSMENT OF SCIENCE CENTER, VALAYA ALONGKORN RAJABHAT UNIVERSITY UNDER ROYAL PATRONAGE

2025-05-19T17:31:52+07:00

The objective of the current study was to assess the organizational carbon footprint of the Science Center at Valaya Alongkorn Rajabhat University under the Royal Patronage. The assessment was carried out based on the guidelines established by Thailand Greenhouse Gas Management Organization (Public Organization). Data collection was conducted for one year between August 2023 to July 2024 which was defined as baseline period for calculating greenhouse-gas emissions in form of carbon dioxide equivalent (CO₂eq). The results were indicated that the total greenhouse-gas emissions of the Science Center were 205,830.40 kilograms of CO₂eq, which were categorized into the following scopes of activities: Scope 1 was equal to 16.60 kgCO₂eq (0.01%). Scope 2 was equal to 173,401.77 kgCO₂eq (84.25%) and Scope 3 was equal to 32,412.58 kgCO₂eq (15.75%). Electricity consumption was identified as the major activity contributing to greenhouse-gas emissions, representing the percentage of 84.25. Then, staff commuting between the university and residences representing the second-largest percentage of 15.50%. The findings were highlighted the necessity to implement effectively energy-saving guidelines and policies within the organization.

OPTIMIZING WORKFORCE UTILIZATION AND LINE BALANCING IN SME TIE-DYE PRODUCTION: AN MPL-WP AND MPL-NWP MODELING APPROACH

2025-07-09T16:43:22+07:00

This study proposed a mathematical model to optimize workforce utilization in Small and Medium-sized Enterprises (SMEs) operating multiple tie-dye production lines involving a total of 10 distinct jobs with processing times ranging from 15 to 60 minutes. Two assignment strategies were developed and compared: the Multiple Production Line with Worker Pool (MPL-WP) model, allowing workers to be flexibly assigned across production lines, and the Multiple Production Line without Worker Pool (MPL-NWP) model, restricting workers to specific lines. The models were formulated as binary integer programming problems incorporating processing time constraints, precedence relations, and takt time to ensure feasible production schedules. Using the Python-MIP optimization package, various scenarios with differing production demands and available times were solved to analyze the relationship among these factors and the required workforce size. Results were indicated that the MPL-WP model generally reduced the total number of workers needed by up to 25% compared to the MPL-NWP model, particularly under high-demand or time-constrained conditions. This has highlighted the operational benefits of worker flexibility in improving labor efficiency and reducing costs. Moreover, the findings were provided actionable insights for SMEs seeking to enhance productivity despite limited resources and fluctuating demands, reinforcing the practicality and relevance of flexible assignment models in real-world production planning.

A DESIGN OF A SOLAR PHOTOVOLTAIC POWER GENERATION SYSTEM AND REAL-TIME PERFORMANCE MONITORING SYSTEM USING IoT TECHNOLOGY

2025-03-14T12:35:34+07:00

This article presented the design of a solar photovoltaic power generation system and real-time performance monitoring system using IoT technology. The primary objective was to design and develop a solar-power generation system equipped with a real-time monitoring system of energy consumption as well as to evaluate the efficiency of the performance monitoring system. The monitoring system was designed to measure voltage, electrical current and power, including various constraints, for examples; maximum current, installed power capacity, and load percentage. All data would be displayed on a smartphone using the Blynk application, allowing users to monitor the status of system as real time. The system design was also emphasized on a stable operation and accurate measurement of electrical current. The research results were indicated that the average error of electrical current measurement was 0.99%, while the average error of voltage measurement was 0.15%, demonstrating high accuracy. The system was tested in real-situation applications by supplying power to an insect-repellent lighting system used in agriculture. The test results showed that the solar photovoltaic power generation system could provide sufficient energy covering the operational period of the load. Furthermore, the performance monitoring system of electrical-power generation enabled users to verify power-generation capabilities and evaluate capacity of system efficiently. It also helped users to precisely plan an energy consumption based on load demand.

SHEET RUBBER QUALITY CLASSIFICATION USING IMAGE PROCESSING AND THE K-MEANS ALGORITHM

2025-02-06T16:22:22+07:00

The research on sheet rubber quality classification using image processing and K-Means algorithm was aimed to 1) study about the algorithm of classifying quality of sheet rubber using image-processing method 2) classify the quality of sheet rubber by using image-processing method and 3) determine the quality classification efficiency of sheet rubber by image processing method. The research methods were as follows: The first step was data collection, the second step was the analysis of need and processes of system, the third step was to design algorithm for quality classification of sheet rubber using image processing, the fourth step was the system development, the fifth step was about testing the system to correct any defects, the sixth step related to evaluate the working performance and the seventh step related to the summary of research results. Statistics for data analysis were average and percentage.

The research results were revealed that 1) the algorithm for classifying the quality of sheet rubber were as follows: 1.1) To design a method of photographing sheet rubber comprised 3 formats including control light format, room light, and outdoors. 1.2) Pre-processing images – images were cropped for creating dataset employing 150 cropped images of both raw and smoked rubber sheet in form of light patterns with size 100 x 100 pixels. But 100 images were used in the clustering process and only 50 images were used to create test data per type
1.3) To analyze the color brightness level employing clustering with K-Means algorithm. 2) The quality of sheet rubber classified by image processing has been divided into 2 groups: The first group was normal sheet rubber and the moldy sheet rubber. Overall, the evaluation of accuracy was 96.57 percent, of precision was 99.95 percent, of recall was 93.19 percent as well as of F1-Score was 96.45 percent, respectively.

DESIGN AND IMPLEMENTATION OF AN IOT-BASED AUTOMATIC WATER PUMP CONTROL SYSTEM FOR RESIDENTIAL APPLICATIONS

2025-06-30T08:57:45+07:00

Inadequate water pressure and inefficient energy utilization remain critical challenges in residential plumbing, especially in densely populated housing complexes during peak demand intervals. Conventional water pumps react directly to pressure fluctuations, leading to frequent cycling, energy inefficiency, and diminished equipment lifetimes. This paper presents an IoT-based control system for household water pumps, incorporating an ESP32 microcontroller, a pressure transducer transmitter (PTT) sensor, and an ultrasonic water level sensor to tackle these difficulties. The system functions in automated and manual modes using the Blynk mobile application, controlling electric ball valve and water pump operations according to real-time data and user-specified thresholds. A case study involving a single residence validated the system's efficacy in sustaining consistent water pressure, surpassing 12.0 psi, with an average of 19.4 psi and a maximum of 23.4 psi. The PTT sensor exhibited a relative error of less than 1.7%, remaining within the ±2% accuracy threshold. Furthermore, the system diminished electrical consumption by 31.4% relative to traditional functioning. The results underscore the system's potential to enhance energy economy, operational dependability, and remote control capabilities as well as facilitate its incorporation into intelligent and sustainable home water management.

A PROTOTYPE OF MALFUNCTION ALERT SYSTEM FOR OXYGEN-GENERATOR EQUIPMENT BASED ON INTERNET OF THINGS

2025-02-05T16:30:12+07:00

The purposes of the current study were to design, improve, and develop a prototype of alert system for aeration control. The research experiment was carried out at the Crab Bank of Phang Sai Community in Kradangnga Subdistrict, Sathing Phra District, Songkhla Province. The system was tested during three power outages over one-month period. The System Development Life Cycle (S.D.LC) framework was employed to facilitate data collection, analysis and operational design. In addition, the development of aquacultural system was integrated with IoT and NB-IoT technologies for real-time monitoring and control. Then, the prototype system was tested for its functionality. To analyze the data, descriptive statistics, and correlation analysis in the context of information management, including mean and standard deviation were employed. The research findings were found that users’ satisfaction towards the effectiveness of the prototype were as follows: data security was average at 4.40 (= 4.40, S.D. = 0.533), usability was average at 4.39 (= 4.39, S.D. = 0.512), and its effectiveness was average at .39 (= 4.39, S.D. = 0.560). Notably, the system successfully mitigated the impact of power outages through timely alerts. In addition, this alert system has been able to be scaled for various types of aquacultures, for example, shrimp and fish farming, by integrating IoT technology for real-time environmental monitoring.

ENHANCING THE VALUE OF COW DUNG AS AN INGREDIENT FOR PRODUCING INTERLOCKING BLOCKS

2024-10-07T10:29:47+07:00

This study aimed to enhance the value of cow dung residue from dairy farms by using it as an ingredient in the production of interlocking bricks. Cow dung obtained from a waste separator was dried and sieved through a No. 4 sieve. The processed cow dung was then mixed with laterite soil as a partial replacement for the soil. The mixture proportions were determined by weight, with a cement-to-laterite ratio of 1:5, and the laterite soil was replaced with cow dung at levels of 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, and 0.7 parts by weight. The water-to-cement ratio was maintained at 0.5. The mixtures were molded into interlocking brick samples with dimensions of 22.5 × 11.5 × 6 cm and cured under room temperature conditions. Water absorption and density were evaluated at 28 days, while compressive strength was tested at 7, 14, 21, and 28 days. The results indicated that water absorption increased with the addition of cow dung. Based on the 602-2547 standard, a cement-to-laterite-to-cow dung ratio of 1 : 4.8 : 0.2 was classified as a load-bearing interlocking block (suitable for mixing with the highest amount of cow dung). Additionally, a cement-to-laterite-to-cow dung ratio of 1 : 4.4 : 0.6 was classified as non-load-bearing interlocking bricks. These products have potential applications in building construction, residential housing, building decoration, landscaping, and garden edging.

DEVELOPMENT AND EVALUATION OF A SMART GREENHOUSE SYSTEM FOR LETTUCE CULTIVATION WITH REAL-TIME SMARTPHONE MONITORING

2025-04-01T10:08:15+07:00

This research study investigated the control of temperature, relative humidity, and soil moisture in a misting greenhouse system using an Arduino control unit to regulate the operation of electrical devices via a smartphone for growing lettuce. In addition, sensors for measuring temperature, relative humidity, and soil moisture were employed to monitor and automatically control the system, with real-time data accessible through a smartphone. The system utilized a 24Vdc, 6.9 bar high-pressure water pump with 12 misting nozzles of 0.6 mm and a single 8-inch 30-Watt exhaust fan. All were connected through the Blynk IoT application via Wi-Fi. The experimental results showed that the system could maintain an average temperature of 30.5°C within the greenhouse, reduce the ambient temperature by an average of 3.48°C, and increase relative humidity by up to 5.0%. The average of relative humidity during the test period was 1.05%. Additionally, the system was able to maintain a stable soil moisture level with an average of 27.5%. The study demonstrated that the system could effectively control temperature, humidity, and soil moisture within the optimal range for lettuce cultivation. Due to the measurement of various system parameters, it was found that the sensors used had a margin of error of approximately ±0.58°C for temperature, ±3.42% RH for relative humidity, and ±2.93% for soil moisture. These errors should be taken into consideration when analyzing the experimental results and designing future systems to enhance the accuracy of control.

VALUE ADDITION AND PRODUCT DEVELOPMENT OF FURIKAKE RICE SEASONING POWDER FROM THE LOCAL SILKWORM PUPAE SPECIES OF SURIN PROVINCE

2025-02-14T09:56:15+07:00

The current research was aimed to develop products made from local silkworm pupae, evaluate their nutritional value, assess product changes, and transfer knowledge to the community. Furikake rice seasoning powder was developed with fish-to-silkworm pupae ratios as follows: Formula 1 (100:0), Formula 2 (50:50), Formula 3 (0:100), and Formula 4 (0:150). Sensory testing was conducted using a 9-point hedonic scale. The experiment was carried out employing Randomized Complete Block Design namely, RCBD with 20 trained testers and 50 participants from the local area. The research results showed that the trained assessors preferred rice seasoning powder (Furikake) from silkworm pupae in the aspect of overall preference and Formula 4 was highly accepted with the value of mean score of 8.10 ± 0.73. In terms of nutritional value per 100 grams, the findings indicated the following contents: total energy, moisture, protein, total fat, carbohydrates, dietary fiber, ash and calcium which were 567.99 kcal, 1.01%, 35.06%, 39.83%, 17.32%, 7.11%, 6.79% and 310.40 mg, respectively. After 3 months of storage, the results were found as follows: total plate count (TPC), yeast and mold (Y&M), and E. coli were 3,000 CFU/g, less than 10 CFU/g, and less than 3 CFU/g, respectively. All of which were within acceptable standards. Due to the analysis of physical and chemical properties, the water activity (aw) was measured at 0.35, and the color values for brightness (L*), redness (a*), and yellowness (b*) were 47.35, 4.40, and 22.98, respectively. The improvement of the product in terms of color to Tom Yum flavor was resulted in increasing the redness (a*) value and the yellowness (b*) value at 13.36 and 29.10, respectively. Silkworm pupae products have been an interestingly new option of protein source.

FORECASTING THE PARTICULATE MATTER 2.5: A CASE STUDY IN CHALOEM PHRA KIAT DISTRICT, SARABURI PROVINCE, THAILAND

2025-05-14T15:34:10+07:00

The airborne particulate matter, which has been especially concerned about the case of PM2.5, has been a prominent factor heavily affecting everyday lives of Chaloem Phra Kiat District's residents. It has negatively influenced the health of the residents, especially as it has become evident in allergic-related skin conditions among the residents. Having understood the decline in the quality of life because of PM2.5 pollution, this study has sought to make predictions for the amounts of PM2.5 concentrations based on the previous data over the last 4 months, 6 months, and 1 year. The forecasting models used in the analysis revolved around the Auto-Regressive Integrated Moving Average (ARIMA), Vector Auto Regression (VAR), and Long Short-Term Memory (LSTM). The Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) were used to estimate the accuracy and performance of these models employing MATLAB and Orange as the data analysis software. It was found that the error rates between the two models of ARIMA and VAR were comparable, while the LSTM model showed significantly lower error rates, with the lowest MAE of 1.67 µg/m³ and MAPE of 7.94%, which was indicative of a better capacity to forecast. Additionally, the study showed clear seasonal fluctuations in the PM2.5 concentrations, which grew steadily to peak during the winter, then fell in summer, and finally fell to their lowest during the rainy season. For example, the peak monthly average in January reached over 55 µg/m³, while in August, it dropped below 15 µg/m³. A consistent cyclical pattern was found every year. As a benchmark forecasting and comparative analysis, this research laid a foundation for further research studies, possibly using advanced machine learning algorithms for further improvement of predictive accuracy and robustness of the models involved.

IMAGE FEATURE EXTRACTION BY DEEP LEARNING MODELS FOR REVERSE IMAGE SEARCH

2025-05-15T11:34:31+07:00

Image feature extraction by deep learning models for reverse image search was aimed to develop a model of the image feature extraction employing deep learning models and evaluate the precise results of reverse image search. This research utilized ResNet50 model, which has been pre-trained transfer learning then well-tuned as the feature extractor for a dataset of 20 Thai-food-image categories that have been internationally popular. The processes mentioned were constructed as a dataset representing a semantic image for comparing with search images using cosine similarity measurement. These processes enabled the fast and accurate image retrieval without the need for labeled data. The research resulted showed the precise evaluation of the reverse image search, especially for the first three results achieving an 80-percent precision. When increasing the number of retrieved results to 5 and 10 images, the precision was at a good level. These results aligned with users’ behavior, who have typically focused only on the top-ranked results. However, similarity among visually alike images, variations in viewpoint, scale, illumination, including background clutter affected errors in recognizing distinctive features. The outcome of this research was an evaluation of the performance of the model employing in the real-situation scenarios, which could serve as a guideline for developing image retrieval systems in e-commerce or duplicating image identification for online media.

DEVELOPMENT OF A LEGAL-CONSULTANT WEB APPLICATION FOR COMMUNITY ENTERPRISES

2025-03-19T12:39:34+07:00

Community enterprises in Thailand have normally faced with challenges in accessing legal knowledge and consultation essentially for proper business operations. This has resulted in entrepreneurs lacking an understanding of legal requirements and potentially made errors in complying with relevant regulations. This research was aimed to develop, test, and evaluate the efficiency of a web application for legal consultation specifically designed for community enterprises. This was also aimed to create an easily accessible tool that provided legal knowledge and consultation to community enterprise entrepreneurs. The system development employed the Systems Development Life Cycle (SDLC) approach and was designed to be directly accessible through web browsers without requiring additional software installation, enabling users to access the system from various devices. The performance evaluation utilized the Black Box Testing method conducting by five experts. The evaluation was employed to cover four key aspects: program accuracy and functionality, convenience and ease of use, program capability, and system security.

Research findings indicated that the developed web application efficiently provided legal knowledge and consultation for community enterprises. The overall evaluation results were at the highest level ( = 4.62, S.D. = 0.44). When examining individual aspects, all areas achieved the highest rating level as follows: the program accuracy and functionality gained the scoring highest (= 4.74, S.D. = 0.43), following by convenience and ease of use ( = 4.67, S.D. = 0.44), system security (= 4.52, S.D. = 0.45), and program capability (= 4.51, S.D. = 0.44), respectively. The current web application served as a valuable tool for community enterprises and could function as a model for developing legal consultation systems in other areas or for applying information technology to support future community enterprise operations.

PHOTOSYNTHETIC-LIGHT RESPONSE CURVE OF FOUR HEMP (Cannabis sativa L.) VARIETIES AT VEGETATIVE STAGE

2025-03-25T11:03:35+07:00

Light energy is an important driving force of photosynthesis in plants. The objective of the current experiment was to determine the light requirement for maximal photosynthesis in hemp (Cannabis sativa L.). Researchers examined the response of photosynthesis under thirteen different levels of light intensity (PPF) from 0–2000 mmolPPF m^-2 s^-1 of four hemp varieties including Auto Blues (Sandy), Sour RNA Seedless, Early Remedy, and Sour Suver Haze Seedless] at vegetative stage. The result showed that net photosynthetic rate (A) rapidly increased in response to increasing PPF range of 0–1000 mmolPPF m^-2 s^-1, then slightly increased based on increasing PPF. The maximum gross photosynthetic rates (A_max) among four hemp varieties were in the range of 36.8–44.2 mmolCO₂ m^-2 s^-1. The light saturation points (I_s) where the A reached their maximum ranged between 916.9–1030.2 mmolPPF m^-2 s^-1 and the light compensation points (I_c) ranged between 77.0–99.0 mmolPPF m^-2 s^-1. The maximum photorespiration rates (R_{l max}) were in the range of 7.9–11.6 mmolCO₂ m^-2 s^-1. The light requirement data from this study could be used for light management in hemp cultivation.