특허 문헌의 유사성 평가 및 검색 연구는 특허 문헌의 효율적인 관리뿐 아니라 산업 및 기술 분야에서효율적이고 빠른 정보 수집을 위해 중요한 주제로 다뤄지고 있다. 특히 특허 도면은 산업 기술의 발전과 혁신의 결과물을 시각적으로 표현해왔으나 지금까지 텍스트에 비해 중요하게 다뤄지지 못한 측면이 있다.
본 연구는 효과적인 특허 도면의 검색을 위해 딥러닝의 대표 모델 ResNet-50과 전통적인 컴퓨터비전알고리즘 SIFT를 이용하여 유사성을 평가하는 연구이다. 먼저 시각적 유형의 유사성을 평가하기 위해총 10,827개의 특허도면을 이용한 유형 분류 실험을 진행했으며 분류성능은 95%가 넘는 Accuracy를나타냈다. 두 번째로 기술도면 5,000개를 사용하여 ResNet과 SIFT를 이용한 검색 실험을 진행하여 각모델의 유사성을 평가하는 특징을 살펴보았다. 마지막으로 원본 데이터 50개와 원본 데이터를 다양한형태로 증강한 데이터 4,800개를 이용하여 편집 유형별로 검색 및 매칭한 결과, ResNet은 72.54%, SIFT는 86.71%의 평균적인 매칭 결과를 나타냈다.
연구 수행 결과, 이미지 전체의 정보를 이용하여 유사도를 비교하는 ResNet-50과 달리 SIFT는 이미지내 특징점 등 속성 정보를 이용하여 유사도를 판단하므로 시각적으로 유사한 이미지를 찾는 일에는ResNet이, 같은 이미지를 찾는 일에는 SIFT가 더 강점이 있는 것으로 평가할 수 있다.
The study of the similarity evaluation and retrieval of patent documents is critical not only for the efficient management of patent literature, but also for the rapid and effective collection of information in industrial and technological fields. Patent drawings visually represent the outcomes of technological advancements and innovations, but have not been given as much importance as texts in the past. This study evaluated the similarity of patent drawings for effective retrieval using the representative deep-learning model, ResNet-50, and the traditional computer vision algorithm, scale-invariant feature transform (SIFT). First, a classification experiment using 10,827 patent drawings was conducted to evaluate the similarity of the visual types, achieving a classification performance with an accuracy exceeding 95%. Second, a retrieval experiment using 5,000 technical drawings was conducted to compare the features of ResNet and SIFT based on their similarity. Finally, the retrieval and matching performances of ResNet and SIFT were evaluated using 50 original data samples and 4,800 augmented data samples created by various forms of editing. ResNet demonstrated an average matching performance of 72.54%, whereas SIFT achieved an average matching performance of 86.71%. The findings reveal that, unlike ResNet-50, which compares similarity using the entire image information, SIFT evaluates similarity based on attribute information, such as key points within the image. Consequently, ResNet is advantageous for identifying visually similar images, whereas SIFT excels in identifying identical images.
The study of the similarity evaluation and retrieval of patent documents is critical not only for the efficient management of patent literature, but also for the rapid and effective collection of information in industrial and technological fields. Patent drawings visually represent the outcomes of technological advancements and innovations, but have not been given as much importance as texts in the past. This study evaluated the similarity of patent drawings for effective retrieval using the representative deep-learning model, ResNet-50, and the traditional computer vision algorithm, scale-invariant feature transform (SIFT). First, a classification experiment using 10,827 patent drawings was conducted to evaluate the similarity of the visual types, achieving a classification performance with an accuracy exceeding 95%. Second, a retrieval experiment using 5,000 technical drawings was conducted to compare the features of ResNet and SIFT based on their similarity. Finally, the retrieval and matching performances of ResNet and SIFT were evaluated using 50 original data samples and 4,800 augmented data samples created by various forms of editing. ResNet demonstrated an average matching performance of 72.54%, whereas SIFT achieved an average matching performance of 86.71%. The findings reveal that, unlike ResNet-50, which compares similarity using the entire image information, SIFT evaluates similarity based on attribute information, such as key points within the image. Consequently, ResNet is advantageous for identifying visually similar images, whereas SIFT excels in identifying identical images.