1、 Standard Practice for Quantifying Cracks in Asphalt Pavement Surfaces from Collected Pavement Images Utilizing Automated Methods AASHTO Designation: PP 67-161 Release: Group 1 (April 2016) American Association of State Highway and Transportation Officials 444 North Capitol Street N.W., Suite 249 Wa
2、shington, D.C. 20001 TS-5a PP 67-1 AASHTO Standard Practice for Quantifying Cracks in Asphalt Pavement Surfaces from Collected Pavement Images Utilizing Automated Methods AASHTO Designation: PP 67-161Release: Group 1 (April 2016) 1. SCOPE1.1. This practice outlines the procedures for quantifying cra
3、cking distress at the network level inasphalt pavement surfaces utilizing automated methods. Detailed specifications are not includedfor equipment, instruments, or the associated software used to process the images. According tothese specifications, any equipment that can be adequately validated to
4、meet the functionalitystipulated herein is considered acceptable. The goal is to achieve a significant level ofstandardization, which will contribute to the production of consistent pavement conditionestimates while not unduly limiting innovation.1.2. The automation level covered by this standard in
5、volves minimal human intervention in theprocess. It is understood that the current level of technology requires significant human review ofthe automation process to verify performance of the system. It is anticipated that the required levelof human intervention will decline over the next few years.1
6、.3. Collect the pavement images to be processed according to PP 68.1.4. Sampling of pavement images from the data collection standard process is acceptable beforeproceeding with the analysis process in this standard. The data sample is to be large enough toprovide the confidence level required by th
7、e agency. Sample size and spacing will depend onconstruction practices and numerous other factors that impact pavement continuity. A 100 percentsample is recommended until a sampling method can be supported by statistical analysis.1.5. This practice does not purport to address all of the safety conc
8、erns, if any, associated with its use.It is the responsibility of the user of this standard to establish appropriate safety and healthpractices and determine the applicability of regulatory limitations related to and prior to its use.2. REFERENCED DOCUMENTS2.1. AASHTO Standards: PP 68, Collecting Im
9、ages of Pavement Surfaces for Distress Detection3. TERMINOLOGY3.1. cracka fissure of the pavement material at the surface with minimum dimensions of 1-mm(0.04-in.) width and 25-mm (1-in.) length. 2016 by the American Association of State Highway and Transportation Officials. All rights reserved. Dup
10、lication is a violation of applicable law.TS-5a PP 67-2 AASHTO 3.2. crack lengththe length measured along the crack path using all available data points between crack termini. 3.3. crack orientationthe angular measurement in degrees between the direction of travel and a line drawn between the crack
11、termini as measured within the measurement zone of interest. 3.4. crack positionthe coordinates of the midpoint of the crack measured perpendicular to the shoulder edge of the pavement and the longitudinal location relative to the starting collection point. 3.5. crack terminusthe point at which the
12、crack width goes below and remains less than 1 mm (0.04 in.) for a 10-mm (0.4-in.) length, or the intersection with another crack, or the end of a summary section. 3.6. crack widththe average gap in millimeters (inches) between the two edges of a crack measured at points along the crack with a spaci
13、ng between 1 mm (0.04 in.) and 10 mm (0.4 in.). 3.7. extent of crackingthe sum of the lengths in meters (feet) of all cracks in the summary section and zone where appropriate. Note 1For pattern cracking, agencies have the option to report extent by area as the sum of the areas in square meters (squa
14、re feet) of all pattern areas for each zone within the summary section. 3.8. inside wheel patha longitudinal strip of pavement 1.0 m (39 in.) wide and centered 0.875 m (35 in.) to the left of the centerline of the lane in the direction of travel. 3.9. lanethe traveled surface between the inside edge
15、 of the pavement markings. In the absence of markings, an equivalent portion of the pavement surface. Note 2Where cracks (or the paved surface edge) exist beyond the pavement markings, but within 0.3 m (1 ft) of the pavement markings, agencies have the option to move the lane edges so that the later
16、al extent of the lane is increased to include the new boundary. 3.10. longitudinal cracka crack at least 0.3 m (12 in.) long and with a crack orientation between and including +20 and 20 degrees relative to the lane centerline. 3.11. measurement zoneone of the five strips of pavement created by the
17、wheel paths and the areas between and outside the wheel paths. 3.12. other crackall cracks not identified as longitudinal, pattern, or transverse. 3.13. outside wheel patha longitudinal strip of pavement 1.0 m (39 in.) wide and centered 0.875 m (35 in.) to the right of the centerline of the lane in
18、the direction of travel. 3.14. pattern cracka crack that is part of a network of cracks that form an identifiable area or grouping of cracks. Note 3An identifiable area or grouping of cracks may be determined by interconnecting cracks with a length of less than 0.3 m (12 in.) or through a process de
19、signed to separate independent cracks from cracks that due to their proximity are likely related. The process could be based on machine learning or other artificial intelligence techniques. One method currently being evaluated establishes 150 mm (6 in.) buffers around each crack and computes how muc
20、h of the buffer area is shared with the buffer areas for other cracks. A crack sharing less than a threshold of its area (50%) with other cracks is considered independent and not part of a pattern. 3.15. pavement image a representation of the pavement that describes a characteristic (gray scale, col
21、or, temperature, elevation, etc.) of a matrix of points (pixels) on the pavement surface. 2016 by the American Association of State Highway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5a PP 67-3 AASHTO 3.16. severity of crackingthe average width
22、 in millimeters (inches) of all cracks in the summary section and zone where appropriate. 3.17. summary sectiona portion of a pavement lane over which the data are summarized. 3.18. transverse cracka crack at least 0.3 m (12 in.) long and with a crack orientation between and including 70 and 110 deg
23、rees relative to the lane centerline. 4. SIGNIFICANCE AND USE 4.1. This practice outlines the procedures for quantifying cracking distress at the network level in asphalt pavement surfaces utilizing automated methods. 4.2. The level of development of the current technology requires that the collecti
24、on and analysis processes be linked together. It is hoped that in the future, a standard set of pavement images may be created by the collection equipment from a standard pavement distress simulation package yet to be developed and used to evaluate and control the performance of this method. 5. DATA
25、 COLLECTION 5.1. Data are collected according to PP 68. 6. DATA REDUCTIONCRACK DETECTION 6.1. The system that detects the presence of a pavement anomaly that is commonly considered a crack typically involves one or more image analysis software routines. 6.2. It is anticipated that future detection p
26、rocesses may involve the analysis of pavement images of multiple characteristics to improve the reliability of the detection process. As of this revision, elevation processes are becoming available. 6.3. The detection process is clearly dependent on the quality of the pavement images to be analyzed.
27、 Currently, there is no reliable method to independently measure the performance of the collection and analysis processes. The criteria listed herein therefore represent the requirements of the total collection and analysis system. It is anticipated that a reliable method will be developed to indepe
28、ndently establish the image quality so that, in the future, each process can be evaluated separately. Until then, a validation process is described in Section 10. 6.4. An acceptable result of this process is a crack map in which at least: 6.4.1. Thirty-three percent of the cracks less than 3 mm (0.1
29、2 in.) and 60 percent of cracks from 3 mm (0.12 in.) to less than 5 mm (0.2 in.) wide are mapped (see Note 4). 6.4.2. Eighty-five percent of the cracks 5 mm (0.2 in.) or wider are mapped (see Note 4). 6.4.3. A crack is considered mapped if at least 85 percent of its length is detected (see Note 4).
30、6.4.4. The pavement images should be sufficiently void of erroneous differences between data point values such that a section of pavement without distress, discontinuities, or pavement markings contains less than 3 m (10 ft) total length of detected false cracking in 50 m2(540 ft2) of pavement 2016
31、by the American Association of State Highway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5a PP 67-4 AASHTO (see Note 4). The determination of this capability will be made utilizing a minimum of ten 0.03-km (100-ft) samples of various pavement ty
32、pes that meet the criteria. Note 4These performance values are the estimates of a panel of experts based on current technology. Ongoing research and equipment developments will better define and improve these values over the next few years. As capabilities are better defined, separate levels of perf
33、ormance may be established for two or three classes of equipment and software. 7. DATA ANALYSISCRACK CLASSIFICATION AND VALUATION 7.1. The summary section distance for cracking data is recommended to be 0.015 km (0.01 mi) or less to more clearly identify extent. 7.2. Detected pattern cracks and long
34、itudinal cracks (not transverse) are separated into five measurement zones across the pavement (see Figure 1). 7.2.1. Zone 1 is between the inside wheel path and the lane edge at the adjacent lane. 7.2.2. Zone 2 is the inside wheel path. 7.2.3. Zone 3 is the space between the wheel paths. 7.2.4. Zon
35、e 4 is the outside wheel path. 7.2.5. Zone 5 is between the outside wheel path and the outside lane edge. Figure 1Cross Section of Survey Lane Showing Wheel Paths and Defined Survey Area between Wheel Paths 7.3. Pattern cracks will be summarized by: 7.3.1. The sum of the lengths in meters (feet) of
36、all pattern cracks in the summary section shall be the extent of pattern cracking for each zone within the summary section. (Additionally, an extent by area can be reported as defined in Note 1 of Section 3.7.) 7.3.2. The average width in millimeters (inches) of all pattern cracks in the summary sec
37、tion for a given zone shall be the severity of the pattern cracking for that zone within the summary section. 2016 by the American Association of State Highway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5a PP 67-5 AASHTO 7.4. Longitudinal crack
38、s will be summarized by: 7.4.1. The sum of the lengths in meters (feet) of all longitudinal cracks in the summary section shall be the extent of longitudinal cracking for each zone within the summary section. 7.4.2. The average width in millimeters (inches) of all longitudinal cracks within the summ
39、ary section for a given zone shall be the severity of longitudinal cracking for that zone within the summary section. 7.5. Other cracks will be summarized by: 7.5.1. The sum of the lengths in meters (feet) of all other cracks in the summary section shall be the extent of other cracking for each zone
40、 within the summary section. The average width in millimeters (inches) of all other cracks within the summary section for a given zone shall be the severity of other cracking for that zone within the summary section. 7.6. Transverse cracks will be summarized by: 7.6.1. The sum of the lengths in mete
41、rs (feet) of all the transverse cracks in the summary section shall be the extent of transverse cracking within the summary section. 7.6.2. The average width in millimeters (inches) of all transverse cracks within the summary section shall be the severity of transverse cracking within the summary se
42、ction. Note 5Cracks detected from intensity images alone and that are greater than 25 mm (1 in.) wide may reflect that a narrower crack is present and was sealed. Intensity images alone cannot separate these conditions. If this is the only technology employed, it is recommended that the system flag
43、these occurrences for technician analysis. This analysis can be supported by input from the data collection operator by flagging when crack sealing is present. The advent of elevation images that can be analyzed along with intensity images should resolve this issue. 8. DATA REPORTING 8.1. The calcul
44、ated extent of longitudinal cracks, pattern cracks, and other cracks shall be reported by measurement zone for each crack category. The calculated extent of transverse cracks shall be reported as a single value for the entire summary section. 8.2. The calculated severity of longitudinal cracks, patt
45、ern cracks, and other cracks in each measurement zone shall be the reported severity for that crack category. The calculated severity of transverse cracks shall be reported as a single value for the entire summary section. Note 6Therefore, each summary section will have 32 values that characterize t
46、he cracking contained therein if both extent and severity are deemed necessary. 9. DATA INTERPRETATION 9.1. The agency is free to utilize the reported data as best fits its pavement management needs. 9.2. Any increase in the extent and severity of cracking in Zones 2 and 4 above that in Zone 3 is ex
47、pected to typically reflect the impact of traffic loading. 9.3. Agencies are alerted that dividing the scalar extent and severity values into level categories or bins can result in erratic results. When values are near the bin limits, natural variation in the data will cause dramatic shifts in resul
48、ts. Such converting can be useful, however, for general comparison to previously collected data in which bin limits were used. 2016 by the American Association of State Highway and Transportation Officials. All rights reserved. Duplication is a violation of applicable law.TS-5a PP 67-6 AASHTO 10. SY
49、STEM VALIDATION 10.1. The process of calibrating and checking the performance of the crack detection and analysis process is left to the agency. Generally, the agency should follow the manufacturers recommendations for calibrating and verifying the performance of the process. The following considerations should be included in any program. 10.2. The interrelationship between the supplied pavement images and the performance of the processing system is evident. 10.3. A standard set of pavement images from the collection equipment should be created that chal
