Table of Contents
A Survey on Correlation Filters for Object Detection: Today, computer vision has become an integrated part of people’s lives. It has, and continues to influence many aspects of daily life and has allowed better working in robotics, real time videos and can also be used for image enhancement, object detection etc. Image processing is one of the rapidly growing technologies and is one of the important parts of computer vision.
Read here: A Survey on Environmental Hazard Monitoring
Of these, image enhancement or the object detection method has become the recently important task due to its many possible applications, such as object recognition, face recognition and video co-segmentation.
However, regardless of the ongoing growth in the field of object detection, it is still a challenging job as the process of detecting object is affected not only by internal factors like accuracy, efficiency, but also by external factors, such as noises, illuminations and occlusions.
Several ways has been introduced to detect the object that is associated with the image or the video. One of the important methods is by using correlation. So, a survey was done among different proposals and this paper consists survey among different methods for object detection.
A Survey on Correlation Filters for Object Detection: Object detection is a prominent task related to image processing that deals with identifying occurrences of objects in a particular classes such as humans, buildings or cars in digital images and videos.
At present, it has become a relevant task and researchers are finding the faster methods for accurately detecting the object. Every object class has its own special features that help in classifying the class. Several methods had been introduced in order for detection of an object. But the procedures and the algorithms used in each method are different.
The main concern is that, how to efficiently and accurately detects the object without causing any delays. In views of that, handling the noises, illuminations, background clutters and occlusions becomes more important. So that the filters must be designed in order to increase the accuracy, efficiency, speed and performance thereby reducing the human time.
There are many popular real world applications that are directly or indirectly related to object detection. With the availability of large amounts of data, faster GPUs, and better algorithms, we can now easily train computers to detect and classify multiple objects within an image with high accuracy. Object detection through automated systems is useful when we need to label an object that gets resided in an image or a video more efficiently.
“A Survey on Correlation Filters for Object Detection”: In this paper latest works in the field of object detection were discussed. Many researchers had contributed and are still working in this field.
Though there are a number of problems in existing systems that need to be addressed such as occlusions, noises and background clutters that are associated with the object.
However some of the problems like redundant images on various scales, detecting different objects on same image have already been solved but still have room for improvement.
It is concluded, correlation filters are introduced to overcome all the existing disadvantages by relating the pixels like difference between foreground and background image, illumination variations etc., thereby it can accurately detect the object by working in a linear fashion.
It will have consistent advantages over the state-of-the-arts when applying this technique to several computer vision applications including eye detection, car detection and object tracking. In all-purpose, different object detection approaches and algorithms can be used in order to get effective results in different real life scenarios.