Recently, IP cameras with H.264 codec have been developed; however the cost of those IP cameras is much higher than that of IP cameras with MJPEG codec and we consider that the high compression rate Sorafenib Tosylate 475207-59-1 offered by H.264 is not necessary for the smoke detection tasks, because it is not necessary to store and/or transmit the captured video sequences between the IP camera modules and the main computer systems. Also MJPEG codec offers higher quality of frames than H.264 codec. Therefore we decided that an MJPEG based IP camera module is the most adequate platform for efficient smoke detection scheme considering computational and economical cost, as well as the frame quality. Although the proposed scheme is designed for MJPEG codec system, it can be adapted to H.264 with minor modifications.
The block diagram of Inhibitors,Modulators,Libraries the proposed smoke detection scheme is shown in Figure 1, which is composed of four stages: video frames acquisition stage, DCT inter-transformation based preprocessing stage, smoke region detection stage and region analysis stage. In the video frames acquisition stage, each frame of size 1,920 �� 1,080 pixels is captured Inhibitors,Modulators,Libraries by an Inhibitors,Modulators,Libraries IP camera and encoded using an standard JPEG codec, in which bi-dimensional DCT is applied to non-overlapped blocks of 8 �� 8 pixels of each frame. In the preprocessing stage, the DCT inter-transformation is applied to all DCT blocks of 8 �� 8 coefficients of each frame to get DCT blocks of 4 �� 4 coefficients without using the inverse DCT (IDCT).
In the smoke region detection stage, using t
Foreground detection algorithms have been implemented in many applications Inhibitors,Modulators,Libraries such as people counting, face recognition, AV-951 license plate detection, crowd monitoring and robotic vision. The accuracy of those applications is heavily dependent on the effectiveness of the foreground detection algorithm used. For example, some people counting systems will not work well when the surrounding illumination is low, such as during rainy days or inside dark rooms. Such a system will not be able to give a correct count because of the inability of the algorithm to distinguish between foreground and background objects. It is very important for the background modelling algorithm to be robust to a variety of complex situations. However, it is almost impossible to make such a system robust to all situations and conditions such as low variation in illumination change, reasonable movement speed and high contrast between background and foreground object. In fact, a majority of previous papers such as [1�C3] only function well within limited conditions and constraints. Any slight deviation from the required conditions significantly degrades performance. Algorithms such as face recognition fail to perform properly Rapamycin AY-22989 once the constraints are violated.