For each pixel of the image obtained by imaging the biological tissue, the image processing apparatus of the present invention, an indication local area including each pixel adjacent to the pixel of interest and the target pixel is or constitutes what shape a feature amount calculation unit first calculates a feature quantity of the first, for each pixel in the image, a feature amount calculation unit of the second calculating a second feature amount indicating a distribution of the gradient direction in the local area an evaluation value calculation unit that calculates for each pixel the shape evaluation value for based on the calculation results of the feature quantity of the second and first feature amounts, it is possible to discriminate the structure of a predetermined shape that is included in the image, and a, and a region extraction unit extracts from the image of candidate region is estimated based on the calculation result of the shape evaluation value, the structure having a predetermined shape is present.本発明の画像処理装置は、生体組織を撮像して得た画像の各画素毎に、注目画素及び注目画素の近傍の各画素を含む局所領域がどのような形状をなしているかの指標を示す第1の特徴量を算出する第1の特徴量算出部と、画像の各画素毎に、局所領域における勾配方向の分布状態を示す第2の特徴量を算出する第2の特徴量算出部と、第1の特徴量及び第2の特徴量の算出結果に基づき、画像に含まれる所定の形状の構造を判別可能とするための形状評価値を各画素毎に算出する評価値算出部と、形状評価値の算出結果に基づき、所定の形状の構造が存在すると推定される候補領域を画像の中から抽出する領域抽出部と、を有する。
