The quality of coffee beans differs from each other based on the geographic locations of its sources. The coffee bean quality is conventionally determined by visual inspection, which is subjective, requiring considerable effort and time and prone to error. This calls for the development of an alternative method that is precise, non-destructive and objective. This paper was conducted with the objective of developing an appropriate computer routine that can characterize coffee beans from the different towns of Cavite, Philippines. Imaging techniques were employed to automatically classify the coffee bean samples according to their specie. Important coffee bean features based in morphology such as area of the bean, perimeter, equivalent diameter, and percentage of roundness were extracted from 195 training images and 60 testing images. Artificial neural network (ANN) and K nearest neighbor (KNN) were employed to automatically categorize the coffee beans. Using ANN, classification scores of 96.66% were achieved while using KNN the following classification scores were achieved 84.12%(k=1), 84.10%(k=2), 81.53%(k=3), 82.56%(k=4), 75.38%(k=5),80.35% (k=6), 38.79%(k=7), 77.44%(k=8), 72.82%(k=9) and 78.45% (k=10). In conclusion, the results of this study have revealed that imaging technique could be used as an effective method to classify coffee bean species. ANN is the more preferred method over KNN in classifying coffee beans.

The quality of a coffee bean is determined by several factors including color, texture, and size. High quality beans are carefully refined where defects, such as black beans, are removed. The assessment through visual inspection may be subjected to external factors such as light and the amount of beans to be inspected. This study presents a method of controlling the coffee bean quality using Image Processing techniques. Normal beans are identified through the extraction of RGB color components of training image. The RGB values were integrated in an image processing technique that eliminated the black beans in the image. Using the technique in this study, a classification of 100% were achieved for eliminating the black beans in the testing images.

This study aimed to design and develop a watermelon ripeness detector using Near-Infrared Spectroscopy (NIRS). The research problem being solved in this study is developing a prototype wherein the watermelon ripeness can be detected without the need to open it. This detector will save customers from buying unripe watermelon and the farmers from harvesting an unripe watermelon. The researchers attempted to use the NIRS technique in determining the ripeness level of watermelon as it is widely used in the agricultural sector with high-speed analysis. The project was composed of Raspberry Pi Zero W as the microprocessor unit connected to input and output devices, such as the NIR spectral sensor and the OLED display. It was programmed by Python 3 IDLE. The detector scanned a total of 200 watermelon samples. These samples were grouped as 60 % for the training dataset, 20 % for testing, and another 20 % for evaluation. The data sets were collected and are subjected to the Support Vector Machine (SVM) algorithm. Overall, experimental results showed that the detector could correctly classify both unripe and ripe watermelons with 92.5 % accuracy.

An Eggplant Fruit and Shoot Borer (EFSB) is a disease that affects the entirety of the eggplant fruit if not detected. Hence, a detector was proposed in the form of a handheld gun. It was designed and developed to non-invasively classify eggplant fruits that are non-infested and infested with EFSB. Using an Arduino Nano as its microcontroller and a near-infrared spectroscopy (NIRS) module, insect infestation is determined and displayed through its OLED display. Measured reflectance data through the NIRS module of the detector is then stored inside a MicroSD module for further use. Since the prototype was developed for online monitoring, portability was given of utmost importance, pattering the design in the form of a handheld gun, inside of which was powered by a 9V rechargeable battery. The 3D-printed chassis of the detector houses the aforementioned components and modules, alongside with switches for power and near-infrared detection. Through Support Vector Machine (SVM), the classifier model was trained and developed using Jupyter and was extracted as a C++ code for the Arduino Nano module. Compared with a farmer's traditional performance in terms of accuracy, precision, and speed, the prototype performed better with an accuracy of 84%, precision of 72.83%, and an average speed of 9.736 seconds.