Spoiled meat classification using semiconductor gas sensors, image processing and neural network

Spoiled meat level can be detected manually by using the senses of sight and smell. However, it can endanger the human body if the gas emitted by rotting meat exhaled directly because of the bacterial contamination. Furthermore, such classifications are inevitably somewhat subjective since everyone has different assessments of the spoiled meat. This research presents the use of semiconductor gas sensors to detect gas emitting from rotting meat as a substitute for human olfaction. In addition, a camera equipped with image processing using Grey Level Co-Occurrence Matrix is applied as a replacement for vision. The responses of gas sensor array and Grey Level Co-occurrence Matrix were processed by Neural Network to classify the spoiled meat level. The classification of Artificial Neural Networks has a high percentage of success up to 82%. This method can replace the role of human senses in meat classification automatically.