Image Edge Extraction For Moving Objects With Spiral Phase

Edge extraction is an important technology in the optical information processing. There are two common types of edge enhancement methods: conventional computer processing and traditional optical system processing. In the computer processing method, the moving objects are extracted from a series of images in the video using a lot of image processing algorithms for edge extraction. Although its control system is simple, it suffers large computation cost,complex calculation process and weak real-time ability. Traditional optical system methods are developed with using high pass filters. This method needs to replace a series of high pass filters in front of a high-speed camera, this approach has the characteristics of obvious detail information of the target and strong real-time capacity. But it is still restricted by environment seriously and the filters vary as the working environment or moving objects varying.The aim of this project "Image edge extraction for moving objects with spiral phase" is to combine the advantages of computer method with optical filtering method. In this topic,video is decomposed into a series of pictures and then by computer simulation a spiral phase virtual optical environment is constructed. Based on the spiral phase theory, the target's edges are extracted. This article provides theoretical references and preparative work for extracting the moving object's edges efficiently and in real time. The main contents of this article can be summarized as follows:First, this paper had analysed the basic theories of edge extraction and applied several traditional algorithms to extract the edge of image. The experiment results suggested that these methods had many defects such as easily effected by noise, restricted by high condition,thresholds setting, error detecting and missing detecting and so on. Because of the radial Hilbert transform theory, a spiral phase filter was located in the Fourier plane in order to distribute the spectrum of the two-dimensional bodies with arbitrary shapes for edge extraction. So in this article a spiral phase virtual optical environment had been constructed based on the radial Hilbert transform theory. After analyzing the theory of spiral phase filer and its output characteristic, not only spiral phase but also spiral phase computed hologram and the computed hologram of Laguerre-Gaussian were loaded on the spatial light monitor(SLM) to acquire the edge information of static objects in the optical experiment. For the existence of the secondary large of traditional spiral phase filter, the modified spiral phase theory was combined with the digital image processing technology, and VC++ (Microsoft Visual C++) was used to write executable files. The process was as follows, the video was decomposed into a series of images. The images were preprocessed. Then the virtual optical environment was constructed by computer simulation. Based on the radial Hilbert transform theory,a high-pass spiral phase filter and a raised-cosine spiral phase filter were respectively located in the Fourier plane in order to distribute the spectrum of the two-dimensional bodies with arbitrary shapes for edge extraction. So far a clearer and continuous edge could be acquired. At last all the final images were composited into a new video only with edge information of objects. This method has many advantages such as non-contact, simple measurement system, high measurement speed, automatically controlled by computer. The experimental results confirmed that the edge information could be acquired effectively by using this method. So this technique can be widely applied in military, industry and medicine fields.