-
公开(公告)号：KR101353952B1
公开(公告)日：2014-01-23
申请号：KR1020120110959
申请日：2012-10-05
Applicant: 계명대학교 산학협력단
CPC classification number: G06K9/6262 , G06K9/00335 , G06K9/00664
Abstract: The present invention relates to a method for detecting a forest fire using spatiotemporal bag-of-features (BoF) and a random forest and, more particularly, to a method for detecting a forest fire using spatiotemporal bag-of-features and a random forest comprising the steps of: (1) whenever a frame of a video sequence is inputted, detecting the difference between the input frame and a previous frame and, if the difference value exceeds a predetermined first threshold, setting the input frame to a key frame; (2) detecting a moving block from the set key frame; (3) extracting a candidate smoke block from the moving block using a smoke color model; (4) generating BoF from the detected candidate smoke block; and (5) performing learning by a random forest with respect to the generated BoF to determine whether the smoke of the candidate smoke block is real. The method proposed by the present invention can set the key frame from the video sequence, extract the candidate smoke block using the non-parametric smoke color model, extract HOG and HOF from the extracted candidate smoke block to generate BoF as spatiotemporal features from the HOG and the HOF, perform learning by the random forest with respect to the generated BoF, thereby enhancing the capability of detecting a forest fire in real time, reducing a false alarm, and accurately detecting smoke caused by the forest fire. [Reference numerals] (AA) Start; (BB) End; (S100) Divide frames forming a video sequence into a plurality of blocks, respectively; (S200) Whenever a frame of the video sequence is inputted, detect the difference between the input frame and a previous frame and, if the difference value exceeds a predetermined first threshold, set the input frame to a key frame; (S300) Detect a moving block from the set key frame; (S400) Extract a candidate smoke block from the moving block using a smoke color model; (S500) Generate bag-of features (BoF) from the detected candidate smoke block; (S600) Perform learning by a random forest with respect to the generated BoF to determine whether the smoke of the candidate smoke block is real
Abstract translation: 本发明涉及使用时空特征（BoF）和随机森林来检测森林火灾的方法，更具体地说，涉及使用时空特征包和随机森林来检测森林火灾的方法 包括以下步骤：（1）每当视频序列的帧被输入时，检测输入帧与先前帧之间的差异，如果差值超过预定的第一阈值，则将输入帧设置为关键帧; （2）从设定的关键帧检测移动块; （3）使用烟色模型从运动块提取候选烟雾块; （4）从检测到的候选烟块产生BoF; 和（5）通过随机森林对生成的BoF执行学习以确定候选烟雾块的烟雾是否是真实的。 本发明提出的方法可以从视频序列设置关键帧，使用非参数烟色模型提取候选烟雾块，从提取的候选烟雾块中提取HOG和HOF，从而生成BoF作为来自HOG的时空特征 和HOF相对于生成的BoF执行随机森林的学习，从而提高实时检测森林火灾的能力，减少误报，准确检测森林火灾造成的烟雾。 （附图标记）（AA）开始; （BB）结束; （S100）分别将形成视频序列的帧划分成多个块; （S200）每当视频序列的帧被输入时，检测输入帧与前一帧之间的差异，并且如果差值超过预定的第一阈值，则将输入帧设置为关键帧; （S300）从设定的关键帧检测移动块; （S400）使用烟色模型从移动块提取候选烟雾块; （S500）从检测到的候选烟雾块产生袋状特征（BoF）; （S600）通过随机森林对生成的BoF执行学习，以确定候选烟雾块的烟雾是否是真实的
-
公开(公告)号：KR101150754B1
公开(公告)日：2012-06-14
申请号：KR1020110032655
申请日：2011-04-08
Applicant: 계명대학교 산학협력단 , (주)제이브이엠
Abstract: PURPOSE: A drug box area detecting system and a method thereof are provided to easily confirm a loaded box state and to control the box state. CONSTITUTION: An image information extracting unit(31) detects the outline of a box. The image information extracting unit extracts the image information. An image analyzing unit(33) calculates the image information. A box control unit(35) generates the control signal according to a state of the box. The box control unit operates a loading box.
Abstract translation: 目的：提供一种药盒区域检测系统及其方法，以容易地确认装箱状态并控制盒状态。 构成：图像信息提取单元（31）检测框的轮廓。 图像信息提取单元提取图像信息。 图像分析单元（33）计算图像信息。 盒控制单元（35）根据盒子的状态产生控制信号。 箱控制单元操作装载箱。
-
3.发明公开
公开(公告)号：KR1020130093861A
公开(公告)日：2013-08-23
申请号：KR1020120015070
申请日：2012-02-15
Applicant: 계명대학교 산학협력단
CPC classification number: G06T7/12 , G06T5/002 , G06T7/168 , G06T2207/10132 , G06T2207/30101
Abstract: PURPOSE: A method of automatically dividing lumen boundaries in an intravascular ultrasound image using a nonparametric probability model and a smoothing function improves the accuracy of division by extracting only a point of interest. CONSTITUTION: An intravascular ultrasound image is converted into a polar coordinate image (S100). A catheter domain is removed in the converted polar coordinate image (S200). An initial point of interest is extracted from the whole polar coordinate image (S300). A point of interest at lumen boundaries is extracted by filtering the initial point of interest (S400). [Reference numerals] (AA) Start; (BB) End; (S100) Step of converting an intravascular ultrasound image obtained by an intravascular ultrasound image device into a polar coordinate image; (S200) Step of removing a catheter domain in the converted polar coordinate image; (S300) Step of extracting an initial point of interest from the whole polar coordinate image without the catheter domain for finding the position of lumen boundaries; (S400) Step of extracting a point of interest at the lumen boundaries by filtering the extracted initial point of interest; (S500) Step of detecting the lumen boundaries in a continuous curve shape from the extracted discrete point of interest at the lumen boundaries; (S600) Step of reversely converting the polar coordinate image for the detected lumen boundaries into the original coordinate image
Abstract translation: 目的：使用非参数概率模型和平滑功能自动分配血管内超声图像中的管腔边界的方法通过仅提取兴趣点来提高分割的准确性。 构成：将血管内超声图像转换为极坐标图像（S100）。 在转换的极坐标图像中移除导管结构域（S200）。 从整个极坐标图像中提取初始感兴趣点（S300）。 通过过滤初始兴趣点来提取流明边界处的兴趣点（S400）。 （附图标记）（AA）开始; （BB）结束; （S100）将通过血管内超声图像装置获得的血管内超声图像转换为极坐标图像的步骤; （S200）在转换的极坐标图像中去除导管结构域的步骤; （S300）从没有导管结构域的整个极坐标图像中提取初始感兴趣点的步骤，以找到流明边界的位置; （S400）通过对所提取的初始兴趣点进行滤波来提取流明边界处的兴趣点的步骤; （S500）从在所述管腔边界处提取的离散感兴趣点检测连续曲线形状的管腔边界的步骤; （S600）将检测到的流明界面的极坐标图像逆变换为原始坐标图像的步骤
-
4.发明授权
公开(公告)号：KR101366341B1
公开(公告)日：2014-02-24
申请号：KR1020120015070
申请日：2012-02-15
Applicant: 계명대학교 산학협력단
Abstract: 본 발명은 비모수적 확률 모델과 스무딩 함수를 이용한 혈관 내 초음파 영상에서 내강 경계면 자동 분할 방법에 관한 것으로서, 보다 구체적으로는 (1) 혈관 내 초음파 영상 장치를 이용하여 획득된 혈관 내 초음파 영상을 극좌표 영상으로 변환하는 단계, (2) 상기 변환된 극좌표 영상에서 도뇨관(catheter) 영역을 제거하는 단계, (3) 상기 도뇨관 영역을 제거한 상기 극좌표 영상 전체로부터 내강 경계면의 위치를 찾기 위한 초기 관심 점을 추출하는 단계, (4) 추출된 상기 초기 관심 점을 필터링하여 내강 경계면 관심 점을 추출하는 단계, 및 (5) 추출된 불연속적인 상기 내강 경계면 관심 점으로부터 연속적인 곡선 형태의 내강 경계면을 검출하는 단계를 포함하는 것을 그 구성상의 특징으로 한다.
본 발명에서 제안하고 있는 비모수적 확률 모델과 스무딩 함수를 이용한 혈관 내 초음파 영상에서 내강 경계면 자동 분할 방법에 따르면, 혈관 내 초음파 영상을 극좌표로 변환한 후 웨이블릿 변환을 적용하여 초기 관심 점을 추출하고, 비모수적 확률 모델과 스무딩 함수를 이용하여 추출된 관심 점 중에서 잡음과 칼슘에 의해 발생되는 점을 제거하면서 내강 경계면에 해당하는 관심 점만을 추출하여 내강 경계면을 분할함으로써 분할 정확도를 향상시켜 내강 경계면을 정확하게 검출할 수 있다.
-
-
-
Search Results
4
records
(took 0.02 seconds)
