This paper proposes the water level measuring method based on the image,while the ruler used to indicate the water level is stained.The contamination of the ruler weakens or eliminates many features which are required...This paper proposes the water level measuring method based on the image,while the ruler used to indicate the water level is stained.The contamination of the ruler weakens or eliminates many features which are required for the image processing.However,the feature of the color difference between the ruler and the water surface are firmer on the environmental change compare to the other features.As the color differences are embossed,only the region of the ruler is limited to eliminate the noise,and the average image is produced by using several continuous frames.A histogram is then produced based on the height axis of the produced intensity average image.Local peaks and local valleys are detected,and the section between the peak and valley which have the greatest change is looked for.The valley point at this very moment is used to detect the water level.The detected water level is then converted to the actual water level by using the mapping table.The proposed method is compared to the ultrasonic based method to evaluate its accuracy and efficiency on the various contaminated environments.展开更多
This paper proposes a night-time vehicle detection method using variable Haar-like feature.The specific features of front vehicle cannot be obtained in road image at night-time because of light reflection and ambient ...This paper proposes a night-time vehicle detection method using variable Haar-like feature.The specific features of front vehicle cannot be obtained in road image at night-time because of light reflection and ambient light,and it is also difficult to define optimal brightness and color of rear lamp according to road conditions.In comparison,the difference of vehicle region and road surface is more robust for road illumination environment.Thus,we select the candidates of vehicles by analysing the difference,and verify the candidates using those brightness and complexity to detect vehicle correctly.The feature of brightness difference is detected using variable horizontal Haar-like mask according to vehicle size in the location of image.And the region occurring rapid change is selected as the candidate.The proposed method is evaluated by testing on the various real road conditions.展开更多
This paper proposes an algorithm that extracts features of back side of the vehicle and detects the front vehicle in real-time by local feature tracking of vehicle in the continuous images.The features in back side of...This paper proposes an algorithm that extracts features of back side of the vehicle and detects the front vehicle in real-time by local feature tracking of vehicle in the continuous images.The features in back side of the vehicle are vertical and horizontal edges,shadow and symmetry.By comparing local features using the fixed window size,the features in the continuous images are tracked.A robust and fast Haarlike mask is used for detecting vertical and horizontal edges,and shadow is extracted by histogram equalization,and the sliding window method is used to compare both side templates of the detected candidates for extracting symmetry.The features for tracking are vertical edges,and histogram is used to compare location of the peak and magnitude of the edges.The method using local feature tracking in the continuous images is more robust for detecting vehicle than the method using single image,and the proposed algorithm is evaluated by continuous images obtained on the expressway and downtown.And it can be performed on real-time through applying it to the embedded system.展开更多
In this paper,detection of a vehicle from a road image with fog is focused to detect an vehicle from a foggy image.Because of the fog in the image,a shape of an object is vague.Therefore an obstacle may occur on the v...In this paper,detection of a vehicle from a road image with fog is focused to detect an vehicle from a foggy image.Because of the fog in the image,a shape of an object is vague.Therefore an obstacle may occur on the vehicle detection.Thus,features from a foggy road image are surveyed through experiments,and a histogram is calculated with the bright value.The stretching method is then applied with the specific weight as the centre to detect a vehicle smoothly.If the high density area,from the view point of histogram,is applied with the stretching method,the definition of the image can be increased.On this fact,this paper proposed a method to divide the histogram and to determine applicable range of the stretching method.The improved results by the proposed methods are proved with the comparison tests between the proposed and previous methods.展开更多
Vehicle detection in still images is a comparatively difficult task.This paper presents a method for this task by using boosted local pattern detector constructed from two local features including Haar-like and orient...Vehicle detection in still images is a comparatively difficult task.This paper presents a method for this task by using boosted local pattern detector constructed from two local features including Haar-like and oriented gradient features.The whole process is composed of three stages.In the first stage,local appearance features of vehicles and non-vehicle objects are extracted.Haar-like and oriented gradient features are extracted separately in this stage as local features.In the second stage,Adaboost algorithm is used to select the most discriminative features as weak detectors from the two local feature sets,and a strong local pattern detector is built by the weighted combination of these selected weak detectors.Finally,vehicle detection can be performed in still images by using the boosted strong local feature detector.Experiment results show that the local pattern detector constructed in this way combines the advantages of Haar-like and oriented gradient features,and can achieve better detection results than the detector by using single Haar-like features.展开更多
The facial expression recognition system using the Adaboost based on the Split Rectangle feature is proposed in this paper.This system provides more various features in increasing speed and accuracy than the Haar-like...The facial expression recognition system using the Adaboost based on the Split Rectangle feature is proposed in this paper.This system provides more various features in increasing speed and accuracy than the Haar-like feature of Viola,which is commonly used for the Adaboost training algorithm.The Split Rectangle feature uses the mask-like shape composed with 2 independent rectangles,instead of using mask-like shape of Haar-like feature,which is composed of 2~4 adhered rectangles of Viola.Split Rectangle feature has less diverged operation than the Haar-like feature.It also requires less operation because the sum of pixels requires only two rectangles.Split Rectangle feature provides various and fast features to the Adaboost,which produces the strong classifier with increased accuracy and speed.In the experiment,the system had 5.92 ms performance speed and 84%~94% accuracy by learning 5 facial expressions,neutral,happiness,sadness,anger and surprise with the use of the Adaboost based on the Split Rectangle feature.展开更多
基金supported by the Brain Korea 21 Project in 2010,the MKE(The Ministry of Knowledge Economy,Korea)the ITRC(Information Technology Research Center)support program(NIPA-2010-(C1090-1021-0010))
文摘This paper proposes the water level measuring method based on the image,while the ruler used to indicate the water level is stained.The contamination of the ruler weakens or eliminates many features which are required for the image processing.However,the feature of the color difference between the ruler and the water surface are firmer on the environmental change compare to the other features.As the color differences are embossed,only the region of the ruler is limited to eliminate the noise,and the average image is produced by using several continuous frames.A histogram is then produced based on the height axis of the produced intensity average image.Local peaks and local valleys are detected,and the section between the peak and valley which have the greatest change is looked for.The valley point at this very moment is used to detect the water level.The detected water level is then converted to the actual water level by using the mapping table.The proposed method is compared to the ultrasonic based method to evaluate its accuracy and efficiency on the various contaminated environments.
基金supported by the MKE(The Ministry of Knowledge Economy),Korea,under the ITRC(Infor mation Technology Research Center)support program supervised by the NIPA(National IT Industry Promotion Agency)(NIPA-2011-C1090-1121-0010)by the Brain Korea 21 Project in2011
文摘This paper proposes a night-time vehicle detection method using variable Haar-like feature.The specific features of front vehicle cannot be obtained in road image at night-time because of light reflection and ambient light,and it is also difficult to define optimal brightness and color of rear lamp according to road conditions.In comparison,the difference of vehicle region and road surface is more robust for road illumination environment.Thus,we select the candidates of vehicles by analysing the difference,and verify the candidates using those brightness and complexity to detect vehicle correctly.The feature of brightness difference is detected using variable horizontal Haar-like mask according to vehicle size in the location of image.And the region occurring rapid change is selected as the candidate.The proposed method is evaluated by testing on the various real road conditions.
基金supported by the Brain Korea 21 Project in 2011 and MKE(The Ministry of Knowledge Economy),Korea,under the ITRC(Infor mation Technology Research Center)support program supervised by the NIPA(National IT Industry Promotion Agency)(NIPA-2011-C1090-1121-0010)
文摘This paper proposes an algorithm that extracts features of back side of the vehicle and detects the front vehicle in real-time by local feature tracking of vehicle in the continuous images.The features in back side of the vehicle are vertical and horizontal edges,shadow and symmetry.By comparing local features using the fixed window size,the features in the continuous images are tracked.A robust and fast Haarlike mask is used for detecting vertical and horizontal edges,and shadow is extracted by histogram equalization,and the sliding window method is used to compare both side templates of the detected candidates for extracting symmetry.The features for tracking are vertical edges,and histogram is used to compare location of the peak and magnitude of the edges.The method using local feature tracking in the continuous images is more robust for detecting vehicle than the method using single image,and the proposed algorithm is evaluated by continuous images obtained on the expressway and downtown.And it can be performed on real-time through applying it to the embedded system.
基金supported by the MKE(The Ministry of Knowledge Economy),Koreathe ITRC(Information Technology Research Center)support program(NIPA-2010-(C1090-1021-0010))the Brain Korea 21 Project in 2010
文摘In this paper,detection of a vehicle from a road image with fog is focused to detect an vehicle from a foggy image.Because of the fog in the image,a shape of an object is vague.Therefore an obstacle may occur on the vehicle detection.Thus,features from a foggy road image are surveyed through experiments,and a histogram is calculated with the bright value.The stretching method is then applied with the specific weight as the centre to detect a vehicle smoothly.If the high density area,from the view point of histogram,is applied with the stretching method,the definition of the image can be increased.On this fact,this paper proposed a method to divide the histogram and to determine applicable range of the stretching method.The improved results by the proposed methods are proved with the comparison tests between the proposed and previous methods.
基金supported by the Korea Research Foundation Grant funded by the Korean Government(MOEHRD),the MKE(The Ministry of Knowledge Economy,Korea)the ITRC(Information Technology Research Center)support program(NIPA-2009-(C1090-0902-0007))
文摘Vehicle detection in still images is a comparatively difficult task.This paper presents a method for this task by using boosted local pattern detector constructed from two local features including Haar-like and oriented gradient features.The whole process is composed of three stages.In the first stage,local appearance features of vehicles and non-vehicle objects are extracted.Haar-like and oriented gradient features are extracted separately in this stage as local features.In the second stage,Adaboost algorithm is used to select the most discriminative features as weak detectors from the two local feature sets,and a strong local pattern detector is built by the weighted combination of these selected weak detectors.Finally,vehicle detection can be performed in still images by using the boosted strong local feature detector.Experiment results show that the local pattern detector constructed in this way combines the advantages of Haar-like and oriented gradient features,and can achieve better detection results than the detector by using single Haar-like features.
基金supported by the Brain Korea 21 Project in2010,the MKE(The Ministry of Knowledge Economy),Koreathe ITRC(Information Technology Research Center)support programsupervised by the NIPA(National ITIndustry Promotion Agency)(NI-PA-2010-(C1090-1021-0010))
文摘The facial expression recognition system using the Adaboost based on the Split Rectangle feature is proposed in this paper.This system provides more various features in increasing speed and accuracy than the Haar-like feature of Viola,which is commonly used for the Adaboost training algorithm.The Split Rectangle feature uses the mask-like shape composed with 2 independent rectangles,instead of using mask-like shape of Haar-like feature,which is composed of 2~4 adhered rectangles of Viola.Split Rectangle feature has less diverged operation than the Haar-like feature.It also requires less operation because the sum of pixels requires only two rectangles.Split Rectangle feature provides various and fast features to the Adaboost,which produces the strong classifier with increased accuracy and speed.In the experiment,the system had 5.92 ms performance speed and 84%~94% accuracy by learning 5 facial expressions,neutral,happiness,sadness,anger and surprise with the use of the Adaboost based on the Split Rectangle feature.