The use of node analysis and logarithmic imagers can improve video analytic applications in the Internet of Things (IoT). The video analytic application attempts to utilize the rich information resources in the daily world for several reasons. Including the daily monitoring of facial recognition hong kong, but most of the reasons are concentrated on predictive analysis and behaviour analysis. The information collected in these applications can be processed more and more extensively through cloud computing. However, depth processing has its limitations, and can be improved in many ways by adding node analysis and logarithmic imagers to the combination.
 

Combination Of Logarithmic Imaging And Nodal Analysis

By adding node analysis to the portfolio and reducing communication with the cloud, data analysis can be improved. The bandwidth requirement of cloud computing is two (if not three) orders of magnitude more than that of node analysis applications. Therefore, the computing power required for node analysis is lower and the delay can be reduced. Densely populated markets, areas with chaotic traffic, and urban parking lots are places with intricate environments. Node analysis can be used for detection for prediction and behaviour analysis. Advanced processing of these environments in the cloud can help formulate business strategies, channel traffic flow, and improve the efficiency of government-managed parking lots. However, using low-end software at sensor nodes instead of performing cloud analysis can improve the latency, bandwidth, security, and power consumption of these scenarios.

In addition to achieving intelligence at the nodes, adding a logarithmic imager to the combination also has advantages for areas where traditional imagers do not meet the requirements, thereby enhancing system functions. In addition to reducing the dependence on brightness changes, the logarithmic imager also provides a higher dynamic range of image processing. For example: in shadows, reflections, sudden light changes and high-contrast scenes, the performance of the logarithmic imager is better than that of the traditional imager. In video applications, the solution of these problems is conducive to data capture, thereby enhancing the node’s analysis capabilities. By improving data capture capabilities, the entire video analytic application can be significantly improved.

The improvement of node analysis technology and logarithmic imager can help solve the video analytic application problem in the Internet of Things. Security, decision-making latency, data bandwidth, and computing power are some of the common engineering problems in IoT applications. These engineering problems can be greatly reduced by reducing data transmission, which is why node analysis is attractive for IoT applications. In video analytic applications, limited contrast and brightness dependence are problems that need to be solved together. The logarithmic imager is the key to video analytic applications and can almost solve this problem. In general, the use of node analysis technology and logarithmic imagers can enhance video analytic applications in the Internet of Things.
 

Intelligent Edge

By processing data based on expected visual events, measurement data can be quickly converted into appropriate actions, without having to transmit any data to the cloud server, or transmit a small amount of data. Quickly analyse the video data instead of sending it to the cloud, and make decisions locally, thereby improving the latency of the system. By reducing the transmission of data that has the risk of interception, not only can the decision-making delay be significantly reduced, but also the security can be improved.

Only the most valuable information needs to be transmitted to the cloud outside the node for prediction or behaviour processing. Optimized data partitioning can give full play to the value of the cloud, because full bandwidth video analytic frames are usually not required. On fixed cameras, most of the visible data between frames is static data, which can be filtered at nodes. The edge node video analytic can provide multiple filtering interpretations to distinguish various expected object types: cars, trucks, bicycles, humans and animals, etc. This extraction operation reduces the data bandwidth and related computing power required on the cloud server, and if the full frame rate video data sent downstream is to be analysed, a large amount of data bandwidth and computing power will be consumed. Compared with cloud computing applications, this reduction in bandwidth can achieve improvements of two or three orders of magnitude, and this is a key performance improvement achieved by node analysis technology.
 

Logarithmic Imaging

By replacing it with logarithmic imaging technology, common problems related to traditional imagers can be solved, thereby further improving video analytic applications. Most conventional imagers are linear imagers, which use the voltage generated by the pixels as a linear function of light, and such pixels result in limited contrast. The linear imager also uses a uniform exposure phase, which limits its dynamic range relative to the pixel exposure time within the frame rate range. Finally, the contrast of traditional imagers depends on brightness, which may cause reflection-related contrast problems. These common problems can be eliminated by replacing it with a logarithmic imager, which uses the voltage generated by the pixel as a logarithmic function of light.

Some traditional imagers are trying to solve the contrast-related issues that prevent users from fully capturing their target environment data. These contrast problems stem from the linear nature of the voltage generated in each pixel. The voltage generated in a linear imaging pixel is proportional to the number of photons irradiated; therefore, its dynamic range is limited compared to logarithmic imaging. Reducing the contrast associated with these linear imagers means reducing the dynamic range. The reduced contrast will adversely affect the analysis of sensor nodes in IoT applications, and ultimately affect the overall performance of the system. The logarithmic imager provides a wider range of brightness levels, thereby increasing the contrast generated by the pixel voltage generated by the logarithm. However, the increase in contrast results in higher photosensitivity, which may not be the desired effect in some applications. Alternatively, increased sensitivity may also be an advantage, depending on the application.

The reflections produced in sunny or bright environments may further hinder the use of traditional imagers for video capture. For example: when there are reflections on the windshield, facial recognition inside the car will become increasingly difficult. This kind of video capture obstacle will introduce errors into the system or lose important data, which will adversely affect video analytic. Since the contrast between pixels of the linear imager depends on the brightness, reflections are generated; therefore, the reflections of the linear imager are more prominent. This dependence on brightness is shown in Equation 1. In addition, the contrast of a logarithmic imager has nothing to do with brightness due to its logarithmic characteristics, which helps to reduce reflections or sudden changes in light.
 

R&D Beyond Individual Components

In order to provide platform-level solutions, ADI began to go beyond the research and development of individual components; these solutions can help customers quickly deploy proven smart solutions to achieve higher performance at lower system costs. Smart applications start with accurate and reliable data, which can be obtained through ADI’s advanced detection and measurement functions. In addition, ADI also cooperates with customers to jointly develop unique system-level solutions that can solve all problems. ADIS1700x is one of the solutions, it can achieve a quarter of the video graphics array (QVGA) imaging analysis.

ADIS1700x is a small size and logarithmic sensitivity QVGA analytical imager module with digital signal processing functions that can optimize video performance. In addition to accelerometers for image stabilization, tilt and impact detection, the module also uses a low-power Blackfin processor for node analysis. In addition, it also uses built-in edge detection technology to track and calculate object motion. Unlike traditional imagers, each 14 μm × 14 μm pixel of a logarithmic imager has a unique exposure phase. Conformal coatings for outdoor operations make them ideal for large-scale deployments, enabling the creation of emerging smart city and building applications. The ADIS17001 is equipped with a 110°field of view (FOV) lens, while the ADIS17002 is equipped with a 67°FOV lens. These two options are suitable for a variety of target applications, including parking lot monitoring, parking violation enforcement, traffic length detection, and industrial analysis.

In general, the use of node analysis technology and logarithmic imagers can significantly improve video applications in the field of Internet of Things. This is also the method that ADI launched together with the ADIS1700x module products. Node analysis rather than cloud computing is conducive to the development of IoT applications. The logarithmic imager has an advantage that its competing products cannot match, and can further improve the application of the Internet of Things. In short, the video analytic application in the Internet of Things is combined with node analysis technology and logarithmic imagers to form a robust system-level solution.