Research On Feature Points Optimized And Background Compression In Visual Imaging

The three-dimensional coordinate measurement is an important part of advanced manufacturing technology. With the refinement of manufacturing as a pillar industry of the national economy, aerospace, aviation, military and civil aviation sectors, large and giant, the product exterior design more and more complex, the scene of large-size precision measurement requirements are getting higher and higher. In the process of calculating the three-dimensional coordinates, the image needs to be obtained first. The quality of the acquired image directly affects the accuracy of the calculated coordinates. Now the camera used by visual measurement obtained the images can meet the basic measurement. But the huge information leading to occupy a large space and slow the transmission speed is the main problem. Currently the occupancy of image could be reduced by using image compression.But due to the deletion of the image information the three-dimensional coordinates will lead to a large error. So the solution that can solve the above problem needs to be proposed.In the visual measurement the characteristics of image was studied and combined with the needs of practical application. Select the aircraft image contains feature coding points to process. Using different compression algorithms between the feature point area and the background area is proposed in the paper to reduce the occupation of the image, while not affecting the information of the feature point.During the experiment, the image was processed by lossless compression in the feature point area and lossy compression in the background area. In order to processing the feature point area, identify the area was needed first. In the paper the image segmentation and contour tracking was used to identify the feature points. Then adjusting the gray grade of the feature point and removing the noise point were used to optimize the image to obtain the image including the complete and accurate feature point. After that, the image was to compressed by the above algorithm. Finally, for testing experimental result, the method in this paper compared with JPEG and TIFF compression to illustrate the characteristics of the algorithm to verify the feasibility of the method.