• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 신호처리소사이어티 추계학술대회 논문집
• /
• pp.453-456
• /
• 2003

Compared with other features of the image, color features are less sensitive to noise and background complication. Besides, this adding to object segmentation has more accuracy of image retrieval. This paper presents object segmentation and HAQ(Histogram Analysis and Quantization) algorithm approach to extract features(the object information and the characteristic colors) of an image. The empirical results shows that this method presents exactly spatial and color information of an image as image retrieval's feature.

• 대한전자공학회논문지SP
• /
• 제41권4호
• /
• pp.113-120
• /
• 2004

본 논문은 내용 기반 검색 기법에 의한 보다 효율적인 특징 추출 및 영상 검색 알고리즘을 제안하였다. 먼저, MPEG 비디오의 key frame을 입력 영상으로 하여 Gaussian edge detector를 이용하여 객체를 추출하고, 그에 따른 객체 특징들, location feature distributed dimension feature와 invariant moments feature를 추출하였다. 다음, 제안하는 HAQ (Histogram Analysis and Quantization) 알고리즘으로 characteristic color feature를 추출하였다. 마지막으로 key frame이 아닌 shot frame을 질의영상으로 하여 제안된 matching 기법에 따라 4가지 특징들의 단계별 검색을 수행하였다. 본 논문의 목적은 사용자가 요구하는 장면이 속한 비디오의 shot 경계 내의 key frame을 검색하는 새로운 내용 기반 검색 알고리즘을 제안함에 있다. 제안된 알고리즘을 바탕으로 10개의 뮤직비디오, 836개의 시험 영상으로 실험한 결과, 효과적인 검색 효율을 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권6호
• /
• pp.3144-3164
• /
• 2019

Face recognition systems have several potential applications, such as security and biometric access control. Ongoing research is focused to develop a robust face recognition algorithm that can mimic the human vision system. Face pose, non-uniform illuminations, and low-resolution are main factors that influence the performance of face recognition algorithms. This paper proposes a novel method to handle the aforementioned aspects. Proposed face recognition algorithm initially uses 68 points to locate a face in the input image and later partially uses the PCA to extract mean image. Meanwhile, the AdaBoost and the LDA are used to extract face features. In final stage, classic nearest centre classifier is used for face classification. Proposed method outperforms recent state-of-the-art face recognition algorithms by producing high recognition rate and yields much lower error rate for a very challenging situation, such as when only frontal ($0^{\circ}$) face sample is available in gallery and seven poses ($0^{\circ}$, ${\pm}30^{\circ}$, ${\pm}35^{\circ}$, and ${\pm}45^{\circ}$) as a probe on the LFW and the CMU Multi-PIE databases.

• ETRI Journal
• /
• 제42권4호
• /
• pp.619-632
• /
• 2020

Highly dispersive S-boxes are desirable in cryptosystems as nonlinear confusion sublayers for resisting modern attacks. For a near optimal cryptosystem resistant to modern cryptanalysis, a highly nonlinear and low differential probability (DP) value is required. We propose a method based on a piecewise linear chaotic map (PWLCM) with optimization conditions. Thus, the linear propagation of information in a cryptosystem appearing as a high DP during differential cryptanalysis of an S-box is minimized. While mapping from the chaotic trajectory to integer domain, a randomness test is performed that justifies the nonlinear behavior of the highly dispersive and nonlinear chaotic S-box. The proposed scheme is vetted using well-established cryptographic performance criteria. The proposed S-box meets the cryptographic performance criteria and further minimizes the differential propagation justified by the low DP value. The suitability of the proposed S-box is also tested using an image encryption algorithm. Results show that the proposed S-box as a confusion component entails a high level of security and improves resistance against all known attacks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권6호
• /
• pp.2910-2925
• /
• 2017

Inadequate and inappropriate intake of food is known to cause various health issues and diseases. Due to lack of concise information about healthy diet, people have to rely on medicines instead of taking preventive measures in food intake. Due to diversity in food components and large number of dietary sources, it is challenging to perform real-time selection of diet patterns that must fulfill one's nutrition needs. Particularly, selection of proper diet is critical for patients suffering from various diseases. In this article, we highlight the issue of selection of proper diet that must fulfill patients' nutrition requirements. To address this issue, we present a cloud based food recommendation system, called Diet-Right, for dietary recommendations based on users' pathological reports. The model uses ant colony algorithm to generate optimal food list and recommends suitable foods according to the values of pathological reports. Diet-Right can play a vital role in controlling various diseases. The experimental results show that compared to single node execution, the convergence time of parallel execution on cloud is approximately 12 times lower. Moreover, adequate accuracy is attainable by increasing the number of ants.

• Archives of Plastic Surgery
• /
• 제49권5호
• /
• pp.614-616
• /
• 2022

As of November 2021, TikTok has one billion monthly active users and is recognized as the most engaging social media platform. TikTok has seen a surge in users and content creators, ranging from athletes to medical professionals. In the past year, content creators have utilized the app to advocate for social reforms, education, and other uses that were not previously considered. Breast cancer is the most commonly diagnosed cancer in women, with an expected 281,550 new cases of invasive breast cancer in 2021. As more individuals with breast cancer choose to undergo resection, the demand for autologous fat grafting in breast reconstruction has increased due to the natural look and feel of breast tissue. The purpose of this article is to analyze content related to breast reconstruction with fat grafting found on TikTok and recommend methods to improve patient education, care, and outcomes. We searched TikTok on November 1, 2021, for videos using the phrase "breast reconstruction with fat grafting." The top 200 videos retrieved from the TikTok search algorithm were analyzed, and all commentaries, duplicates, and nonrelevant videos were removed. Video characteristics were collected, and two independent reviewers generated a DISCERN score A total of 131 videos were included in the study. They were found to have a combined 1,871,980 likes, 41,113 comments, and 58,662 shares. The videos had an average DISCERN score of 2.16. Content creators had an overall low DISCERN score in items involving the use of references, disclosure of risks for not obtaining treatment, and support for shared decision-making. When stratified, the DISCERN score was higher for videos created by physicians (DISCERN average 2.48) than for videos created by nonphysicians (DISCERN average 1.99; p < 0.001).