FRUIT RECOGNITION USING MACHINE LEARNING

dc.contributor.authorNgo, Ana
dc.date.accessioned2021-10-27T13:22:20Z
dc.date.available2021-10-27T13:22:20Z
dc.date.created2020
dc.description2020 Celebration of Student Research and Creativity presentationen_US
dc.description.abstract"Object recognition is really helpful in detecting objects, such as human, cars, buildings… in an image or a video. In this research, we use some machine learning techniques to recognize different kinds of fruits in an image. This research requires software platforms (Python 3, Pycharm, Anaconda) and libraries in Python (numpy, panda, scikit-learn…) for coding part. After taking 30 photos of five common fruits (apple, orange, banana, grape, strawberry), we use Selective Search Algorithm to locate potential objects in those images by drawing bounding boxes. To better capture the characteristics of different fruits, we employ the Histogram of Oriented Gradients (HOG) feature to represent each potential fruit object. This feature is used to train a Support Vector Machine (SVM) model to classify those bounding boxes into different categories. This fruit recognition system works well on most of the collected images. However, the system recognition accuracy decreased to recognize some similar fruits, like apple and strawberry. The reason for this is that these fruits share similar color and shape information. For the future, we will improve recognition accuracy by using other machine learning or deep learning models. It is also possible for us to build a phone application to help people recognize different fruits in the real world."en_US
dc.description.urihttps://youtu.be/WgCWBWJvCw4en_US
dc.identifier.urihttp://hdl.handle.net/11216/4072
dc.language.isoen_USen_US
dc.publisherNorthern Kentucky Universityen_US
dc.relation.ispartofseriesCelebration of Student Research and Creativity;2020
dc.subjectMachine learningen_US
dc.subjectRecognition (Psychology)en_US
dc.subjectFruiten_US
dc.titleFRUIT RECOGNITION USING MACHINE LEARNINGen_US
dc.typePresentationen_US

Files