Computer Vision Technique and Its Application on Moving Object Tracking

Computer Vision Technology Computer vision develops the theory and algorithmic foundation that can automatically extract and analyze useful information on the world from observation images, image sets, or image sequences in special purpose or general purpose computer calculations To science. (Note) The main purpose of computer vision applications is to create an automatic recognition system that can outperform human beings or ultimately achieve superior performance. Computer vision can be used to enable new relational technologies and to connect physical and virtual worlds.

I have always been fascinated with computer vision and its applications. When I started business four years ago, I started using Motion & Tracking software. Spot Angels uses computer vision to create maps. I personally excited about the potential use of all the image data we collect from the car. Automatic driving vehicles are the best way to learn and understand current trends. I decided to join the first automated driving car nanodegree @ Udacity queue. I am planning to follow up on this 9-month course in my free time. My goal is to learn how to manufacture automatic driving cars at the end of the project.

Computer vision is the computational process of artificial vision technology. Based on years of attempts to mimic the human visual system, computer vision has taken many steps since the first iteration of the 1960s and 1970s. In essence, in order to produce artificial vision ability, we must build a mathematical model that allows us to model correctly. First of all, this begins with the extraction of edges and lines, but then shifts to more complex methods such as polyhedral modeling, image segmentation, image warping.

These may be complicated techniques that deserve some of themselves. There are three general problems in computer vision: (1) detection of objects in images (which can also take into account segmentation or segmentation of images) and (2) classification of objects in images. Another is used to identify your previous lion, classification is an important aspect of computer vision. (3) tracking of objects in the video stream, also known as video tracking. As we can imagine, as autonomous car engineers you should be able to deal with these three: images and videos, things in pedestrians, trucks, other cars, etc.

Move target detection is an important research subject in the field of computer vision and video processing. Detection of a video stream of a moving object is the first relevant step in the extraction of information in many computer vision applications. This paper proposes an improved background subtraction of moving object detection in fixed camera state. The adaptive background subtraction is then combined with the symmetric difference to obtain an integrity foreground image. The chromaticity difference is used to remove the shadow of the moving target, and the moving shadow and the moving target are effectively distinguished. The results showed that the algorithm can quickly establish a background model and quickly detect a complete moving target.