TAILIEUCHUNG - Recent Advances in Signal Processing 2011 Part 6

Tham khảo tài liệu 'recent advances in signal processing 2011 part 6', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | 162 Recent Advances in Signal Processing Image Database Fig. 1. System architecture. Image Acquisition The image database considered in this research work is composed by grayscale images acquired during a pavement surface visual survey over a Portuguese road. A digital camera was manually positioned by the inspector with its optical axis perpendicular to the road surface at a distance of approximately m. Images with different sizes are obtained 2048x1536 pixels and 1858x1384 pixels according to different camera setup procedures. The digital camera is oriented in such a way that the images only contain areas belonging to the road pavement surface. Moreover the database includes images with several types of cracks longitudinal transversal and miscellaneous as well as images without any cracks. Instead of processing the images at a pixel level in all the steps of the proposed system each image is divided into a set of non-overlapping regions of size 75x75 pixels. These dimensions were empirically chosen leading to a faster processing time and lower memory storage requirements while providing a good compromise between complexity and accuracy. Database images can then be represented by smaller matrices where each of their values corresponds to the computation of region local statistics as described next. Supervised Crack Detection and Classification in Images of Road Pavement Flexible Surfaces 163 Selection of Training Images Dealing with supervised classification strategies training data images for the envisaged application is necessary for classifiers learning. This section describes a technique for the automatic selection of images to be included in TIS from the entire image database acquired during the visual road pavement survey. To allow a correct learning stage training images should contain road pavement cracks. Therefore in a preliminary classification phase all images are pre-processed in order to detect the regions with most evident crack pixels .

TÀI LIỆU LIÊN QUAN
69    272    11
TỪ KHÓA LIÊN QUAN
Đã phát hiện trình chặn quảng cáo AdBlock
Trang web này phụ thuộc vào doanh thu từ số lần hiển thị quảng cáo để tồn tại. Vui lòng tắt trình chặn quảng cáo của bạn hoặc tạm dừng tính năng chặn quảng cáo cho trang web này.