Update of pattern recognition

Slides:



Advertisements
Similar presentations
Efforts to Improve the Reconstruction of Non-Prompt Tracks with the SiD Lori Stevens UCSC ILC Simulation Reconstruction Meeting May 15, 2007 Includes contributions.
Advertisements

Analysis of the Stereo Hits and the 2D Circle Fitter Hans Wenzel, Hogan Nguyen Feb 9 th, 2011 Introduction Hans implemented stereo hits, formed by the.
Computing EVO meeting, January 15 th 2013 Status of the Tracking Code Gianluigi Boca, Pavia University.
More on making fake TT clusters More creating Fake TT clusters We can compute the number of combinations instead of the purity (following all the possible.
MICE TPG RECONSTRUCTION Tracking efficiency Olena Voloshyn Geneva University.
TRACK DICTIONARY (UPDATE) RESOLUTION, EFFICIENCY AND L – R AMBIGUITY SOLUTION Claudio Chiri MEG meeting, 21 Jan 2004.
June 6 th, 2011 N. Cartiglia 1 “Measurement of the pp inelastic cross section using pile-up events with the CMS detector” How to use pile-up.
Z Segmentation Update (improving reconstruction of non- prompt tracks with the SiD) Lori Stevens (SLAC/UCSC) August 21, 2007 ILC Simulation Reconstruction.
Tracker Software Update Adam Dobbs, MICE CM37 7 th Nov 2013.
1 Tracking Reconstruction Norman A. Graf SLAC July 19, 2006.
MdcPatRec Tracking Status Zhang Yao, Zhang Xueyao Shandong University.
Status of TPC experiment ---- Online & Offline M. Niiyama H. Fujimura D.S. Ahn W.C. Chang.
1 Th.Naumann, DESY Zeuthen, HERMES Tracking meeting, Tracking with the Silicon Detectors Th.Naumann H1 DESY Zeuthen A short collection of experiences.
Tracking in High Density Environment
1 A first look at the KEK tracker data with G4MICE Malcolm Ellis 2 nd December 2005.
Secondary Vertex reconstruction for the D + Elena Bruna University of Torino ALICE Physics Week Erice, Dec. 6 th 2005.
Fast Tracking of Strip and MAPS Detectors Joachim Gläß Computer Engineering, University of Mannheim Target application is trigger  1. do it fast  2.
Roberto Barbera (Alberto Pulvirenti) University of Catania and INFN ACAT 2003 – Tsukuba – Combined tracking in the ALICE detector.
Muon detection in NA60  Experiment setup and operation principle  Coping with background R.Shahoyan, IST (Lisbon)
Search for High-Mass Resonances in e + e - Jia Liu Madelyne Greene, Lana Muniz, Jane Nachtman Goal for the summer Searching for new particle Z’ --- a massive.
K 0 s reconstruction in ALICE Giuseppe Lo Re INFN, Sezione di Catania, and Dipartimento di Fisica e Astronomia, Università di Catania, Italy, March 2002.
Status report - Tracking code - T.Gogami 9/30/2010.
Confirming L1 decision Use: DaVinci v7r4 field 043 Idea (Teubert’s) : Most of min.bias L1 possitive trigger are due to missmeasurements of Pt (P) How mutch.
1 HBD Update Itzhak Tserruya DC meeting, May 7, 2008 May7, 2008.
STAR SVT Self Alignment V. Perevoztchikov Brookhaven National Laboratory,USA.
VXDBasedReco TRACK RECONSTRUCTION PERFORMANCE STUDIES Bruce Schumm University of California at Santa Cruz ALCPG Workshop, Snowmass Colorado August 14-28,
1 MDC Track Finding at BESIII Zang Shilei BES Annual Meeting (Jun 1, 2005) ______________________________________ -Outline- 
MICE Tracker Software A. Dobbs CM32 9 th Feb 2012.
M. Ellis - MICE Collaboration Meeting - Wednesday 27th October Sci-Fi Tracker Performance Software Status –RF background simulation –Beam simulation.
BESIII offline software group Status of BESIII Event Reconstruction System.
Software Update Takashi HACHIYA RIKEN 2012/2/10RIKEN VTX software meeting1.
Object-Oriented Track Reconstruction in the PHENIX Detector at RHIC Outline The PHENIX Detector Tracking in PHENIX Overview Algorithms Object-Oriented.
Tracking software of the BESIII drift chamber Linghui WU For the BESIII MDC software group.
Software Update Takashi HACHIYA RIKEN 2012/1/31VTX software meeting1.
Ftksim at high luminosity Monthly meeting September 22, 2008 Anton Kapliy.
Torino, June 15 th, 2009 Status of the Pattern Recognition with the Hough transform and the STT system alone. Gianluigi Boca 1.
Frascati September 8 th, 2014 Universita’ di Pavia Italy Improvements in the offline Pattern Recognition in the 20 MHz interaction rate environment Gianluigi.
Comparison of algorithms for hit reconstruction in the DTs: Test of calibration procedures for t trig and drift velocity on Test Beam data Test of calibration.
IPHC, Strasbourg / GSI, Darmstadt
Hough-4-14.
Status of Hough Transform TPC Tracker
New TRD (&TOF) tracking algorithm
Alignment of the ALICE MUON Spectrometer
Relative alignment of LXE and DCH using AIF
neural tracking in ALICE Alberto Pulvirenti – I.N.F.N. Catania
Using IP Chi-Square Probability
More technical description:
Progress on the online tracking algorithm
Beam Gas Vertex – Beam monitor
SiD tracking using VXD as a primary tracking device
Status of the Offline Pattern Recognition Code GSI, December 10th 2012
VTX tracking issues Y. Akiba.
Integration and alignment of ATLAS SCT
Tracker to Solenoid Alignment
Tracking Pattern Recognition
Status of the Offline Pattern Recognition Code Goa, March 12th 2013
External alignment Maya Shimomura (ISU) 12/03/15 RIKEN VTX meeting.
Hough Based Tracking For Low Pt In MDC
Individual Particle Reconstruction
Stereo hit in Hough Tracking
Fake Jet Rejection: Sum Jt
Hellenic Open University
Structure from Motion with Non-linear Least Squares
Fast Track Fitting in the SiD01 Detector
RIKEN VTX software Meeting
Calibration of DT-MTCC data
road – Hough transform- c2
Hough-D(mc)- ππJpsi(Data)
Current Status of the VTX analysis
Structure from Motion with Non-linear Least Squares
Presentation transcript:

Update of pattern recognition 2011/8/25 Ryohji Akimoto

Update of pattern recognition Pattern recognition : grouping hit clusters which are supposed to be from an identical track. Pattern recognition in the current code starts from a cluster on the first sublayer, then search clusters on the next sublayer around projection point calculated by already grouped clusters (or collision vertex). The projection point at the next sublayer is calculated by straight projection. But, if more than 3 clusters are grouped, bending angle can be calculated. makes search area narrow and therefore makes fake track rate decrease.

Quick pT calculation The calculation of bending angle in pattern recognition is enough to estimate roughly but should be done quickly. Circle fit in XY-plane is proper, but iteration is necessary if more than 4 clusters are fitted due to non-linearity of chi-square. There are some ways of circle fit without iteration. Conformal mapping : circle passing the origin transformed to line Riemann fit

Riemann fit transform XY-plane to the Riemann sphere in 3D-space by the following transformation Any points on a circle in XY-plane are transformed both on the Riemann sphere and on a plane in 3D-space. Transformed points are fitted by plane in 3D-space ( ) with the least chi-square method.

Riemann fit v.s. iterative fit 0.3 GeV/c 0.5 GeV/c 1 GeV/c 5 GeV/c 10 GeV/c iterative fit Riemann fit Checked with single track simulation. At low pT (< 1GeV/c), pT reconstructed by Riemann fit is comparable to that by iterative fit. But at high pT (> 5GeV/c), pT reconstructed by Riemann fit is worse than that by iterative fit.

phi - projected phi phi difference : 0.3 GeV/c 0.5 GeV/c 1 GeV/c real hit 5 GeV/c 10 GeV/c projected hit Rhit Rdiff collision vertex phi difference :

Reconstruction efficiency ± 0.05 rad ± 0.1 rad 300 MeV/c 97.66% 99.31% 500 MeV/c 98.81% 99.60% 1 GeV/c 99.25% 99.84% 5 GeV/c 99.53% 99.66% 10 GeV/c 99.57% 99.78% efficiency : (real hit is within the area in phi) / (tracks reconstructed by the current code with more than 4 clusters)

Time consumption Simulation event : generated with HIJING (100 event) Time consumption at svxreco.C with circle projection (cut : 0.05 rad) SvxStandAloneReco::process_event() : 192.98 sec. / 100 events Total : 622.74 sec. / 100 event with line projection (current code) SvxStandAloneReco::process_event() : 300.28 sec. / 100 events Total : 755.43 sec. / 100 event ~20 % improved in total

Reconstructed tracks with circle projection (cut : 0.05 rad) : 475,252 segments / 100 events with line projection (current code) : 140,110 segments / 100 events quality cut : < -5 # of segments with circle projection is larger than that with line projection. outlook : check fake track rate with mixed event comes from “track-related” fake track (radiation or etc..) ?