A fast online incremental learning method for object detection and pose classification using voting and combined appearance modeling

This paper presents a novel and rapid object detection method that identifies object positions and classifies object views. To overcome the limitations of appearance-based object recognition, we integrate a spatial relationship between local key points and object center position. A voting technique is applied to estimate the object area and then construct a bounding box to capture the object. A combined appearance model is introduced by a recall image to help deal with false detection problems. Experimental results show that our method can improve the object detection time while still preserving the average precision results. Moreover, our method can improve the accuracy of view classification.     

Moving object detection and tracking from video captured by moving camera

This paper presents an effective method for the detection and tracking of multiple moving objects from a video sequence captured by a moving camera without additional sensors. Moving object detection is relatively difficult for video captured by a moving camera, since camera motion and object motion are mixed. In the proposed method, the feature points in the frames are found and then classified as belonging to foreground or background features. Next, moving object regions are obtained using an integration scheme based on foreground feature points and foreground regions, which are obtained using an image difference scheme. Then, a compensation scheme based on the motion history of the continuous motion contours obtained from three consecutive frames is applied to increase the regions of moving objects. Moving objects are detected using a refinement scheme and a minimum bounding box. Finally, moving object tracking is achieved using a Kalman filter based on the center of gravity of a moving object region in the minimum bounding box. Experimental results show that the proposed method has good performance.     

Caged mice mating behavior detection in surveillance videos

The purpose of this study is to develop a computer vision-based method to automatically detect the mating behavior of caged mice in surveillance videos. Previously we took advantage of our developed algorithm and analyzed the objects of mating mice in the consecutive frames, we unprecedentedly showed that, to the best of our knowledge, the mice mating behavior can be automatically detected based on video processing (Lo et al., 2009 [13]). In this paper, we proposed an improved method which monitors the distance between two mating objects and more effectively detects the mating behavior. In addition, a more detailed portrayal of the mating behavior can be further elaborated as a function of the distance patterns in the tails of two caged mice. Experimental results show that the current system can effectively detect the mice mating behavior with the highest precision rate of 96.1%, far better than that of our previously proposed method.     

Video object segmentation in rainy situations based on difference scheme with object structure and color analysis

A scheme based on a difference scheme using object structures and color analysis is proposed for video object segmentation in rainy situations. Since shadows and color reflections on the wet ground pose problems for conventional video object segmentation, the proposed method combines the background construction-based video object segmentation and the foreground extraction-based video object segmentation where pixels in both the foreground and background from a video sequence are separated using histogram-based change detection from which the background can be constructed and detection of the initial moving object masks based on a frame difference mask and a background subtraction mask can be further used to obtain coarse object regions. Shadow regions and color-reflection regions on the wet ground are removed from the initial moving object masks via a diamond window mask and color analysis of the moving object. Finally, the boundary of the moving object is refined using connected component labeling and morphological operations. Experimental results show that the proposed method performs well for video object segmentation in rainy situations.     

Handheld object detection and its related event analysis using ratio histogram and mixture of HMMs

This paper proposes a novel system to analyze human-object interaction events happening between hands and faces in real time. Two challenging problems in this event analysis must be addressed, i.e., there is no prior knowledge (like shape, color, size, and texture) about the handheld objects, and there are large spatial–temporal variations in event representation. For the first challenge, a novel ratio histogram is proposed to find important color bins to locate handheld objects and their trajectories via a code book technique. This scheme is different from other boosted methods which require very time-consuming estimations to search reliable body configurations. For the second challenge, a mixture of HMMs is proposed to describe an event not only from its dynamic context but also its multiplicity context. It can be performed in real time because an exhaustive search process is avoided to find possible interaction pairs between objects and body parts.