Abstract: An image-based detecting system for traffic parameters first sets a range of a vehicle lane for monitoring control, and sets an entry detection window and an exit detection window in the vehicle lane. When the entry detection window detects an event of a vehicle passing by using the image information captured at the entry detection window, a plurality of feature points are detected in the entry detection window, and will be tracked hereafter. Then, the feature points belonging to the same vehicle are grouped to obtain at least a location tracking result of single vehicle. When the tracked single vehicle moves to the exit detection window, according to the location tracking result and the time correlation through estimating the information captured at the entry detection window and the exit detection window, at least a traffic parameter is estimated.

Abstract: An object tracking method for a non-overlapping-sensor network works in a sensor network. The method may comprise a training phase and a detection phase. In the training phase, a plurality of sensor information measured by the sensors in the sensor network is used as training samples. At least an entrance/exit is marked out within the measurement range of each sensor. At least three characteristic functions including sensor spatial relation among the sensors in the sensor network, difference of movement time and similarity in appearance, are estimated by an automatically learning method. The at least three characteristic functions are used as the principles for object tracking and relationship linking in the detection phase.

Abstract: An image-based detecting system for traffic parameters first sets a range of a vehicle lane for monitoring control, and sets an entry detection window and an exit detection window in the vehicle lane. When the entry detection window detects an event of a vehicle passing by using the image information captured at the entry detection window, a plurality of feature points are detected in the entry detection window, and will be tracked hereafter. Then, the feature points belonging to the same vehicle are grouped to obtain at least a location tracking result of single vehicle. When the tracked single vehicle moves to the exit detection window, according to the location tracking result and the time correlation through estimating the information captured at the entry detection window and the exit detection window, at least a traffic parameter is estimated.

Abstract: A method for detecting a shadow of an object in an image is provided. A moving object in a plurality of continuous images is detected. A histogram of a color variation of the moving object in each of the images is calculated. The histograms of the color variation are accumulated to obtain a cumulative histogram. A distribution of the color variation in the cumulative histogram is estimated to obtain a shadow distribution function. Whether each pixel in a received image belongs to the shadow is determined by using the shadow distribution function.

Abstract: Disclosed is a system and method of image-based space detection. The system includes an image selection module, a 3-layer detection mechanism and an optimization module. At least one image processing area that may affect space-status judgment is selected from plural image processing areas. The 3-layer detection mechanism having an observation layer, a labeling layer, and a semantic layer observes the information about the selected image processing area, associates with a local classification model, and adjacent local constraint model and a global semantics model to completely describe the probability distribution of the links among the three layers, and provide global label constraint information. The optimization module analyzes the probability distribution and global label constraint information, and generates an image-based optimized space detection result.

Abstract: An object tracking method for a non-overlapping-sensor network works in a sensor network. The method may comprise a training phase and a detection phase. In the training phase, a plurality of sensor information measured by the sensors in the sensor network is used as training samples. At least an entrance/exit is marked out within the measurement range of each sensor. At least three characteristic functions including sensor spatial relation among the sensors in the sensor network, difference of movement time and similarity in appearance, are estimated by an automatically learning method. The at least three characteristic functions are used as the principles for object tracking and relationship linking in the detection phase.

Abstract: A method for detecting a shadow of an object in an image is provided. A moving object in a plurality of continuous images is detected. A histogram of a color variation of the moving object in each of the images is calculated. The histograms of the color variation are accumulated to obtain a cumulative histogram. A distribution of the color variation in the cumulative histogram is estimated to obtain a shadow distribution function. Whether each pixel in a received image belongs to the shadow is determined by using the shadow distribution function.

Abstract: Disclosed is a system and method of image-based space detection. The system includes an image selection module, a 3-layer detection mechanism and an optimization module. At least one image processing area that may affect space-status judgment is selected from plural image processing areas. The 3-layer detection mechanism having an observation layer, a labeling layer, and a semantic layer observes the information about the selected image processing area, associates with a local classification model, and adjacent local constraint model and a global semantics model to completely describe the probability distribution of the links among the three layers, and provide global label constraint information. The optimization module analyzes the probability distribution and global label constraint information, and generates an image-based optimized space detection result.