PROBLEM TO BE SOLVED: To provide an egg classification device, an egg classification method and a computer program capable of classifying eggs more appropriately than before. SOLUTION: An egg classification device has a transmitted light acquisition unit that acquires transmitted light data representing light transmitted through an egg, and teacher data including state information representing an internal state of the egg and transmitted light data about the egg. A deep learning model that outputs state information representing the internal state of the egg when the transmitted light data acquired by the transmitted light acquisition unit is input, and a deep learning model that outputs the parameters are learned using the deep learning model. It is provided with a determination unit that determines the classification destination of eggs according to the state information. [Selection diagram] Fig. 2【課題】卵を従来よりも適切に分類することを可能にする卵分類装置、卵分類方法及びコンピュータプログラムを提供する。【解決手段】卵分類装置は、卵を透過した光を表す透過光データを取得する透過光取得部と、卵の内部の状態を表す状態情報及び前記卵についての透過光データを含む教師データを用いてパラメータが学習されており、前記透過光取得部が取得した透過光データが入力された場合に卵の内部の状態を表す状態情報を出力する深層学習モデルと、該深層学習モデルが出力する状態情報に応じて、卵の分類先を決定する決定部とを備える。【選択図】図２
