|-------------------------------------------------------------------------------------------------------------------| | REGION |SUB|COMMENTS |ISSUES | |-------------------------------------------------------------------------------------------------------------------| | c9_5539r| 0 |Medium res, many goo|JPL tool piled up 90 craters at 6 pixel diameter. All 429 original JPL craters w| | | |d craters, "crater01|ere classified correctly as 0's, but then the 398 surviving (real) craters were | | | |" region. (PROCESSED|later reclassified to 3's in tobill*v2.reg (probable tool problem). tobill*v3.re| | | |IN 2009 REPORT, AND |g fixed this problem, but it is unclear which version (class 0's or 3's) was use| | | |2010) |d as the base for that file. In v3, 282 class 3 craters remain (out of 392 survi| | | | |ving JPL craters, or 72%) which seems extremely high, but the image does appear | | | | |visually to have more class 3's than 4's so the stats may be reasonable after al| | | | |l. 58 of the 282 final type 3 JPL craters had a minimum size of 6 pixels. JPL c| | | | |raters were quantized into 39 unique bins with slightly worse resolution than ha| | | | |nd-marked craters | |c10_2878r| 0 |Medium/high res of A|JPL tool piled up 162 craters at 6 pixel diameter. All 493 original JPL craters | | | |sgard region. (PROCE|were classified correctly as 0's, but then the 463 surviving (real) craters were| | | |SSED IN 2009 REPORT,|later reclassified to 3's in tobill*v1.reg (probable tool problem). tobill*v2.re| | | |AND 2010) |g fixed this problem, but it is unclear which version (class 0's or 3's) was use| | | | |d as the base for that file. In v2, 288 class 3 craters remain (out of 446 survi| | | | |ving JPL craters, or 65%) which seems extremely high (and the image looks like i| | | | |t should have more class 4's than 3's visually). 86 of the 288 final type 3 JPL| | | | |craters had a minimum size of 6 pixels (54 class 4's). JPL craters were quantize| | | | |d into 59 unique bins with slightly better resolution than hand-marked craters | |c30_5378r| 1 |High res, "HiRes_02"|Initially the whole region was marked by the JPL tool, then split into sub-regio| | | |region. (PROCESSED I|ns. JPL tool piled up 54 final craters at 6 pixel diameter. All 793 original JPL| | | |N 2009 REPORT, AND 2|craters (in the whole region) were classified correctly as 0's, but then 211 of | | | |010) |328 craters (64%) in this sub-region were later reclassified to 3's. This seems | | | | |high, but the image visually appears to have more class 3's than 4's so it might| | | | |be OK. 31 of the 211 final type 3 JPL craters had a minimum size of 6 pixels (2| | | | |2 class 4's). JPL craters were quantized into 65 unique bins with slightly bette| | | | |r resolution than hand-marked craters | |c30_5378r| 2 |High res, "HiRes_02"|Initially the whole region was marked by the JPL tool, then split into sub-regio| | | |region. (PROCESSED I|ns. JPL tool piled up 44 final craters at 6 pixel diameter. All 793 original JPL| | | |N 2009 REPORT, AND 2|craters (in the whole region) were classified correctly as 0's, but then 225 of | | | |010) |306 craters (74%) in this sub-region were later reclassified to 3's. This seems | | | | |high, but the image visually appears to have more class 3's than 4's so it might| | | | |be OK. 36 of the 225 final type 3 JPL craters had a minimum size of 6 pixels (8| | | | |class 4's). JPL craters were quantized into 73 unique bins with slightly better | | | | |resolution than hand-marked craters | |c30_5378r| 3 |High res, "HiRes_02"|Initially the whole region was marked by the JPL tool, then split into sub-regio| | | |region. (PROCESSED I|ns. JPL tool piled up 15 final craters at 6 pixel diameter. All 793 original JPL| | | |N 2009 REPORT, AND 2|craters (in the whole region) were classified correctly as 0's, but then 75 of 9| | | |010) |6 craters (78%) in this sub-region were later reclassified to 3's. This seems hi| | | | |gh, but the image visually appears to have more class 3's than 4's so it might b| | | | |e OK. 12 of the 75 final type 3 JPL craters had a minimum size of 6 pixels (3 c| | | | |lass 4's). JPL craters were quantized into 35 unique bins with OK resolution | |c30_5453r| 0 |High res, "HiRes_03"|JPL tool piled up only 7 craters at 6 pixel diameter, so smaller R-plot bin stat| | | |region. (PROCESSED I|s might not be too bad. All 210 original JPL craters were classified correctly a| | | |N 2009 REPORT, AND 2|s 0's, but then all 210 later reclassified to 3's in tobill*v1.reg (probable too| | | |010) |l problem). tobill*v2.reg fixed this problem, but it is unclear which version (c| | | | |lass 0's or 3's) was used as the base for that file. In v2, 129 class 3 craters | | | | |remain (out of 208 surviving JPL craters, or 62%) which seems high, but the imag| | | | |e does appear visually to have more class 3's than 4's (it looks very similar to| | | | |c30_5378r with a similar distribution of 3's and 4's) so the stats may be reason| | | | |able after all. 4 of the 208 final type 3 JPL craters had a minimum size of 6 p| | | | |ixels (3 class 4's). JPL craters were quantized into 46 unique bins with worse r| | | | |esolution than hand-marked craters | | | | | | |c30_5253r| 1 |High res, "HiRes_01"|JPL tool piled up 22 craters at 5 pixel diameter (of 143), so no problems at 6 p| | | |region. ISIS MAP PRO|ixels. 100 unique quantization bins used (with 80 populated), so no problems the| | | |JECTED. PROCESSED FO|re either. | | | |R 2010 REPORT. Combi| | | | |ne sub-region stats | | | | |together. | | |c30_5253r| 2 |High res, "HiRes_01"|JPL tool piled up 9 craters at 5 pixel diameter (of 103), so no problems at 6 pi| | | |region. ISIS MAP PRO|xels. 100 unique quantization bins used (with 54 populated), but only 2 bins bet| | | |JECTED. PROCESSED FO|ween 6-7 pixels (.0516km-.0602km) and 5 between 7-8 (.0602km-.0688km) so some mi| | | |R 2010 REPORT. Combi|nor quantizing might be visible at these crater sizes (but is real)?? | | | |ne sub-region stats | | | | |together. | | |-------------------------------------------------------------------------------------------------------------------|