Gaussian Mixture Model (GMM) is a widely used approach for the background subtraction and the moving objects detection. However, the classical GMM probably detects incorrectly and cannot deal with the shadows with a pixel-level and time-domain classification, and thus it cannot monitor the water surface floats effectively. To solve this problem, an improved GMM-based automatic segmentation method (IGASM) is proposed to detect the water surface floats in this paper, where the background updating strategy is improved to segment the water surface floats more effectively. Besides, the GMM results are mapped into an HSV color space, and a light-shadow discriminant function is applied to solve the problems of light and shadow. Then, a morphological method is used to smooth the extracted foregrounds. Finally, Graph Cuts algorithm is applied to optimized the segmentation results according to the spatial information of video images. Experimental results demonstrate that IGASM can detect the water surface floats quickly and accurately, and the influences of light, shadows and ripples of water surface can be eliminated as much as possible.

/pdf/a-gmm-based-segmentation-method-for-the-detection-of-water-82o9y20thk.pdf

A GMM-Based Segmentation Method for the Detection of Water Surface Floats

Blob detection is a common task in vision-based applications. Most existing algorithms are aimed at execution on general purpose computers; while very few can be adapted to the computing restrictions present in embedded platforms. This paper focuses on the design of an algorithm capable of real-time blob detection that minimizes system memory consumption. The proposed algorithm detects objects in one image scan; it is based on a linked-list data structure tree used to label blobs depending on their shape and node information. An example application showing the results of a blob detection co-processor has been built on a low-powered field programmable gate array hardware as a step towards developing a smart video surveillance system. The detection method is intended for general purpose application. As such, several test cases focused on character recognition are also examined. The results obtained present a fair trade-off between accuracy and memory requirements; and prove the validity of the proposed approach for real-time implementation on resource-constrained computing platforms.

https://www.mdpi.com/1424-8220/16/6/782/pdf

A Linked List-Based Algorithm for Blob Detection on Embedded Vision-Based Sensors

In this paper, we propose an adaptive approach to the detection of moving shadows by employing local neighboring information. The process of the proposed approach is mainly operated by three steps: the first step is to detect the candidate shadows by RGB ratio; the second step is to extract partial accurate shadows in order to estimate accurate threshold parameters of shadow detectors; the final step is to utilize three detectors to detect real shadows from candidate shadows. The main contributions of this paper include two parts: an effective method of candidate shadows detection is presented; an adaptive mechanism of estimating threshold parameters is designed. Moreover, three detectors that consist of color, texture and gradient features are jointly utilized to detect shadows at pixel-level. Experimental results on a benchmark suit of indoor and outdoor video sequences demonstrated the proposed approach’s effectiveness.

Adaptive Moving Shadows Detection Using Local Neighboring Information

In order to solve the moving objects shadow problem in foreground extraction of surveillance video images, a new cast shadow detection algorithm based on the YCbCr color space and topological cutting was proposed. Preliminary shadow removal was first performed based on the difference of the three components in the YCbCr color space of shadows and foregrounds. Besides, the maximum-flow/minimum-cut algorithm for image segmentation was optimized considering the topological constraints. The optimal segmentation of the foreground image was obtained during the continuous updating of the label. Finally, two sets of experiments were performed in video image sequences, including real surveillance videos and a well-known benchmark test set. By comparing with two other existing algorithms, the feasibility and effectiveness of the cast shadow detection algorithm were verified by the smooth border and higher recognition accuracy. In addition, the adaptability to foreground object density and different light intensities was measured in an airport terminal, showing that this algorithm can provide a high quality of moving foreground detection in surveillance video images and can be applied in monitoring of public places.

Cast shadow detection based on the YCbCr color space and topological cuts

A vehicle detection and shadow elimination method based on greyscale information, edge information, and prior knowledge

Shadow area causes false detection of moving objects during segmentation and tracking those objects properly. Shadows are also reason for the texture loss of background and false connectivity of independent blobs. Hence, we propose a simple method to detect moving object's cast shadow and then remove the shadow region from Video frames. We extract the moving object by subtraction algorithm based on the difference of pixels. Texture, variance property and intensity in HSV color space are used to detect the shadow region and shadow removal is based on the information from the reference frame. Background information is initially stored in reference frame. The next incoming frames with object are compared with this frame. Color information for both background subtraction and shadow detection to improve object segmentation are ensured in this paper. Experimental results show that our proposed method is easy to be understood, can detect and remove the shadow and extract the moving object properly.

Animesh Kar

Papers

Moving cast shadow detection and removal from Video based on HSV color space