<![CDATA[Healthy tilapia will experience a decrease in quality when stored out of water even with the refrigerator. The decline in fish quality can be seen from the fish's eye. In this study the aim was to develop a tilapia freshness classification system based on fisheye images utilizing image processing and machine learning. Fisheye images were taken at intervals of 0 to 16 hours after being removed from the water. Ten tilapias were used in this study. The distance between the camera and the fish has also been changed with a variation of 4 distances. The total data obtained is 640 fisheye images. The method used in fisheye image classification consists of two stages, namely feature extraction and classification. The four types of statistical features used from image pixel values are the mean, standard deviation, skewness, and kurtosis. While at the classification stage using K-Nearest Neighbor. The scenario that has been determined is then used at the system implementation stage using the Python Programming Language. Testing and analysis using k-fold cross validation and confusion matrix. In the results of the study, the accuracy rate was obtained with an average of 100% using 2 classes, namely 0-2 hours and 2-4 hours. The accuracy rate using 4 classes was obtained with an average of 75%, namely classes 0-2 hours, 2-4 hours, 10-12 hours, and 14-16 hours, and when using all classes, an average accuracy rate of 45% is obtained.]]>
Copyrights © 2023
