Design of Intelligent Monitoring System Based on Low Cost CMOS Camera

Considering the basic requirements of digital monitoring system and the value, performance, power consumption, volume and development trend of CMOS cameras, an intelligent digital image monitoring system based on low cost CMOS cameras is implemented on ordinary PC by using VC++ and VB programming. When the monitoring scene changes, the system automatically takes photos and videos, provides three alarm levels, implements the corresponding processing mechanism, and makes sound alarm, and if necessary enters the manual intervention.

1 System Functions

The system is composed of a computer, a CMOS camera and software. Its basic functions are as follows:

(1) It can overcome the instability of inexpensive CMOS cameras and the defects of low pixel resolution, and provide reliable and stable monitoring for monitoring scenes.

(2) Intelligence and automation can be achieved. When an anomaly is discovered, different handling mechanisms can be implemented based on the level of risk of the anomaly.

(3) Capture photos and video recordings. Record relevant scenes in case of abnormalities, provide image evidence for future analysis, and provide time information.

(4) It has perfect post-processing ability. In this system, you can view the photos and videos taken, make it easy for users to browse and view, and edit, print and so on.

(5) Flexible switch from automatic monitoring to manual operation.

2 Hardware Design

The system hardware composition diagram is shown in Figure 1.

The system host consists of a common PC loading software. The system is designed to take into account both high performance and low price. It uses a large amount of disk space to store huge data and a stable CPU to ensure long-term operation of the system.

The CMOS camera is Logitech Express 2000. Its pixel value is only 100,000, when the resolution is 320 × At 240 pixels, the FPS is 15 frames and the illumination is between 100-100,000 lux.

3 Software Design

3.1 Main Program Diagram

The main program diagram of the monitoring software is shown in Figure 2.

3.2 Monitor Diagram

The monitor block diagram is shown in Figure 3.

The system sets alerts at three levels, green alert, yellow alert and red alert. From low to high, the security of the monitoring scene, the alert status of the system and the alert status of the system are represented respectively.

When the system is in a green alert state, the video image captured by the CMOS camera is replayed in the video playback window located in the center of the screen. The status bar located below the screen displays information such as the running time of the system and the current time. At the same time, the internal monitor detects motion and triggers high alerts and drives the accessory device to work whenever the scene changes.

4 pairs of low cost CMOS cameras

4.1 Image Acquisition

Logitech Quick-Cam SDK is used to acquire camera images. This is a COM interface driver. By programming its settings, you can get images captured by the camera. Figure 4 shows how the programmer works with the lower driver of the camera.

By calling the WDM driver, the application can communicate with the CMOS camera.

4.2 Pre-treatment Method

Noise can degrade the quality of an image during its generation and transmission. In order to suppress noise and improve image quality, the image must be smoothed.

In order to filter out the noise in the image, a smoothing technique is used in the system design. The smoothing technique used in this system is to use a local operator on the noise image, that is, to operate only on some pixels in its local small domain. This method, also known as unweighted domain averaging, treats every pixel in the domain equally. Set the gray value of a pixel in the image to f(x,y) and its field S to N × N, the total number of points set is M, then the gray value of the smoothed point is the average value of the whole area.

Neighborhood Operation is essentially a template operation, that is, the value of a pixel point is not only related to the gray scale of the pixel, but also to the value of its domain point. The idea of smoothing a template is to filter out noise by averaging the pixel values of a point and eight points around it.

The image template operation functions used in the system are as follows:

With this generic template, local smoothing can be easily achieved.

Research on 4.3 Imaging Rule

The most serious problem with inexpensive CMOS cameras is poor stability. That is, if the same static scene is taken continuously, its images will also be very different. For this reason, the experiment on the stability of continuous still image shooting is carried out first. The image is separated by three primary colors, the difference of each pixel point is compared, and the number of different pixel points is counted.

Suppose a particular point is adjacent t₁,t₂The three primary color values corresponding to the time are (R)₁,G₁, B₁) and (R)₂,G₂, B₂). Experiments found that R₁_R₂,G₁_G₂, B₁_B₂But it has a relationship R₁-R₂Delta G₁-G₂Delta B₁-B₂.

The separation of three primary colors is the basis of separating the background image before separation. For this reason, the constructed functions f(1), f(2) represent the values of the three primary colors function corresponding to a point at the time of T1 and t2, respectively. It is found that when |f(1)-f(2)|=a=45 (a is called domain value), the background image before T1 and T2 can be separated accurately, and the system reflects quickly, which is a very suitable compromise point. Therefore, when a is less than or equal to 45, it can be considered that the three primary color difference values at the time of T1 and T2 in the image are caused by camera errors, otherwise it can be judged that an object has entered the monitoring area.

The procedure for separating the foreground from the background is as follows:

Figure 5 shows the result of background and front separation after the program is run.

5 Conclusion

In the process of system research, our group has invested a lot of experience in the research of low cost CMOS camera, and successfully overcome the problems of poor camera stability and low quality. Explore a low-cost multimedia monitoring system development path.

Through the system test in different places, the system has achieved high and stable real-time monitoring function. However, with the further requirements of the monitoring system, in order to adapt to the new situation, the system needs to be optimized in the following areas.

(1) Automatic switching of video signals, automatically switching the camera monitoring area at a certain interval.

(2) Increase the network function to transmit photos and videos to the remote host when alarming.

(3) Increase the function of telephone alarm so that an alarm call can be made.

(4) Deep study on system stability.

(5) Improve the image and video compression algorithm, using MPEG4 compression technology.