Accurate Moving Target Detection Based on Background Subtraction and SUSAN

Abstract

Moving target detection is a fundamental problem in computer vision. Most target detection methods only detect the rough area of the target. However, there are also many applications (e.g., advanced video surveillance system, hybrid video camera system, etc.) need to detect the accurate target area for analyzing the movement or behavior. These applications always product noisy video frames and will lead to high false rates using previous method. In this paper, a two-step accurate moving target detection method is proposed. This method obtains a rough target area using a color background subtraction method with a feedback to background estimation, followed by a modified SUSAN method to estimate the accurate target edge. The modified SUSAN method uses the grads magnitudes to achieve an adaptive threshold and the gray barycenter criterion for denoising. Experimental results demonstrate the proposed method is effective for accurate moving target detection in different noise level video frames compare to several competitive methods proposed in the literature.

Pendeteksi Posisi Keberadaan Manusia dalam Ruangan Menggunakan Metode Perbedaan Citra dengan Sensor Webcam

“…Zhang dkk. [5] mengusulkan metode deteksi target bergerak dalam frame-frame video yang banyak noisenya. Metode yang diusulkan terdiri dari dua tahap deteksi.…”

High Accuracy Detection and Tracking of Objects

“…Jun Zhang et al [20] have proposed a two-step accurate moving target detection method. Their method obtained a rough target area using a color background subtraction method with a feedback to background estimation, followed by a modified SUSAN method to estimate the accurate target edge.…”

Robust Vision-based Multiple Moving Object Detection and Tracking from Video Sequences

“…In the Optical Flow moving object detection procedure [2], the flow field image is calculated and the distribution of the feature is done by cluster processing which is better. But this procedure is not suitable for real-time processing because of its large amount of calculation and its sensitivity to noise and lack of anti-noise performance [3]. On the other hand, in background subtraction procedure [3], the moving object is detected by subtracting the background from the current frame, is a simple procedure and in the case of already known background, this process could able to provide a complete information about the objects.…”

“…But this procedure is not suitable for real-time processing because of its large amount of calculation and its sensitivity to noise and lack of anti-noise performance [3]. On the other hand, in background subtraction procedure [3], the moving object is detected by subtracting the background from the current frame, is a simple procedure and in the case of already known background, this process could able to provide a complete information about the objects. For the feature extraction, there are many available methods such asSpeed-Up Robust Features (SURF) which is a speed-up Version Scale Invariant Feature Transforms (SIFT).…”