Open-Harbor/osrm-backend
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 3 Wiki Activity

Split trace if timestamp delta is over threshold

Browse Source 
Even when matching is not broken we split the trace, if the
sampling frequency goes below 2 samples/minute.
This commit is contained in:
Patrick Niklaus 2015-03-03 00:38:09 +01:00
parent d8d46e0f3e
commit 0d879ed290
1 changed files with 61 additions and 63 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

120
routing_algorithms/map_matching.hpp
View File

@ -63,7 +63,7 @@ constexpr static const double IMPOSSIBLE_LOG_PROB = -std::numeric_limits<double>
constexpr static const double MINIMAL_LOG_PROB = -std::numeric_limits<double>::max(); constexpr static const double MINIMAL_LOG_PROB = -std::numeric_limits<double>::max();
constexpr static const unsigned INVALID_STATE = std::numeric_limits<unsigned>::max(); constexpr static const unsigned INVALID_STATE = std::numeric_limits<unsigned>::max();
constexpr static const unsigned MAX_BROKEN_STATES = 6; constexpr static const unsigned MAX_BROKEN_STATES = 6;
constexpr static const unsigned MAX_BROKEN_TIME = 180; constexpr static const unsigned MAX_BROKEN_TIME = 30;
} }
// implements a hidden markov model map matching algorithm // implements a hidden markov model map matching algorithm
@ -285,9 +285,7 @@ template <class DataFacadeT> class MapMatching final
} }
} }
void initialize(const HiddenMarkovModel& model, void initialize(const Matching::CandidateLists& candidates_list)
unsigned initial_timestamp,
const Matching::CandidateLists& candidates_list)
{ {
// json logger not enabled // json logger not enabled
if (!logger) if (!logger)
@ -304,16 +302,8 @@ template <class DataFacadeT> class MapMatching final
state.values["coordinate"] = osrm::json::make_array( state.values["coordinate"] = osrm::json::make_array(
candidates_list[t][s].first.location.lat / COORDINATE_PRECISION, candidates_list[t][s].first.location.lat / COORDINATE_PRECISION,
candidates_list[t][s].first.location.lon / COORDINATE_PRECISION); candidates_list[t][s].first.location.lon / COORDINATE_PRECISION);
if (t < initial_timestamp)
{
state.values["viterbi"] = osrm::json::clamp_float(Matching::IMPOSSIBLE_LOG_PROB); state.values["viterbi"] = osrm::json::clamp_float(Matching::IMPOSSIBLE_LOG_PROB);
state.values["pruned"] = 0u; state.values["pruned"] = 0u;
}
else if (t == initial_timestamp)
{
state.values["viterbi"] = osrm::json::clamp_float(model.viterbi[t][s]);
state.values["pruned"] = static_cast<unsigned>(model.pruned[t][s]);
}
timestamps.values.push_back(state); timestamps.values.push_back(state);
} }
states.values.push_back(timestamps); states.values.push_back(timestamps);
@ -350,18 +340,20 @@ template <class DataFacadeT> class MapMatching final
} }
void add_viterbi(const unsigned t, void set_viterbi(const std::vector<std::vector<double>>& viterbi,
const std::vector<double>& current_viterbi, const std::vector<std::vector<bool>>& pruned)
const std::vector<bool>& current_pruned)
{ {
// json logger not enabled // json logger not enabled
if (!logger) if (!logger)
return; return;
for (auto s_prime = 0u; s_prime < current_viterbi.size(); ++s_prime) for (auto t = 0u; t < viterbi.size(); t++)
{ {
osrm::json::get(*object, "states", t, s_prime, "viterbi") = osrm::json::clamp_float(current_viterbi[s_prime]); for (auto s_prime = 0u; s_prime < viterbi[t].size(); ++s_prime)
osrm::json::get(*object, "states", t, s_prime, "pruned") = static_cast<unsigned>(current_pruned[s_prime]); {
osrm::json::get(*object, "states", t, s_prime, "viterbi") = osrm::json::clamp_float(viterbi[t][s_prime]);
osrm::json::get(*object, "states", t, s_prime, "pruned") = static_cast<unsigned>(pruned[t][s_prime]);
}
} }
} }
@ -415,7 +407,7 @@ template <class DataFacadeT> class MapMatching final
} }
DebugInfo debug(osrm::json::Logger::get()); DebugInfo debug(osrm::json::Logger::get());
debug.initialize(model, initial_timestamp, candidates_list); debug.initialize(candidates_list);
unsigned breakage_begin = std::numeric_limits<unsigned>::max(); unsigned breakage_begin = std::numeric_limits<unsigned>::max();
std::vector<unsigned> split_points; std::vector<unsigned> split_points;
@ -424,7 +416,51 @@ template <class DataFacadeT> class MapMatching final
prev_unbroken_timestamps.push_back(initial_timestamp); prev_unbroken_timestamps.push_back(initial_timestamp);
for (auto t = initial_timestamp + 1; t < candidates_list.size(); ++t) for (auto t = initial_timestamp + 1; t < candidates_list.size(); ++t)
{ {
// breakage recover has removed all previous good points
bool trace_split = prev_unbroken_timestamps.size() < 1;
// use temporal information if available to determine a split
if (trace_timestamps.size() > 0)
{
trace_split =
trace_split ||
(trace_timestamps[t] - trace_timestamps[prev_unbroken_timestamps.back()] >
Matching::MAX_BROKEN_TIME);
}
else
{
trace_split = trace_split || (t - prev_unbroken_timestamps.back() >
Matching::MAX_BROKEN_STATES);
}
if (trace_split)
{
unsigned split_index = t;
if (breakage_begin != std::numeric_limits<unsigned>::max())
{
split_index = breakage_begin;
breakage_begin = std::numeric_limits<unsigned>::max();
}
split_points.push_back(split_index);
// note: this preserves everything before split_index
model.clear(split_index);
unsigned new_start = model.initialize(split_index);
// no new start was found -> stop viterbi calculation
if (new_start == Matching::INVALID_STATE)
{
break;
}
prev_unbroken_timestamps.clear();
prev_unbroken_timestamps.push_back(new_start);
// Important: We potentially go back here!
// However since t > new_start >= breakge_begin
// we can only reset trace_coordindates.size() times.
t = new_start+1;
}
unsigned prev_unbroken_timestamp = prev_unbroken_timestamps.back(); unsigned prev_unbroken_timestamp = prev_unbroken_timestamps.back();
const auto &prev_viterbi = model.viterbi[prev_unbroken_timestamp]; const auto &prev_viterbi = model.viterbi[prev_unbroken_timestamp];
const auto &prev_pruned = model.pruned[prev_unbroken_timestamp]; const auto &prev_pruned = model.pruned[prev_unbroken_timestamp];
const auto &prev_unbroken_timestamps_list = candidates_list[prev_unbroken_timestamp]; const auto &prev_unbroken_timestamps_list = candidates_list[prev_unbroken_timestamp];
@ -487,58 +523,18 @@ template <class DataFacadeT> class MapMatching final
} }
} }
debug.add_viterbi(t, current_viterbi, current_pruned);
if (model.breakage[t]) if (model.breakage[t])
{ {
BOOST_ASSERT(prev_unbroken_timestamps.size() > 0);
// save start of breakage -> we need this as split point // save start of breakage -> we need this as split point
if (t < breakage_begin) if (t < breakage_begin)
{ {
breakage_begin = t; breakage_begin = t;
} }
BOOST_ASSERT(prev_unbroken_timestamps.size() > 0);
// remove both ends of the breakage // remove both ends of the breakage
prev_unbroken_timestamps.pop_back(); prev_unbroken_timestamps.pop_back();
bool trace_split = prev_unbroken_timestamps.size() < 1;
// use temporal information to determine a split if available
if (trace_timestamps.size() > 0)
{
trace_split =
trace_split ||
(trace_timestamps[t] - trace_timestamps[prev_unbroken_timestamps.back()] >
Matching::MAX_BROKEN_TIME);
}
else
{
trace_split = trace_split || (t - prev_unbroken_timestamps.back() >
Matching::MAX_BROKEN_STATES);
}
// we reached the beginning of the trace and it is still broken
// -> split the trace here
if (trace_split)
{
split_points.push_back(breakage_begin);
// note: this preserves everything before breakage_begin
model.clear(breakage_begin);
unsigned new_start = model.initialize(breakage_begin);
// no new start was found -> stop viterbi calculation
if (new_start == Matching::INVALID_STATE)
{
break;
}
prev_unbroken_timestamps.clear();
prev_unbroken_timestamps.push_back(new_start);
// Important: We potentially go back here!
// However since t+1 > new_start >= breakge_begin
// we can only reset trace_coordindates.size() times.
t = new_start;
breakage_begin = std::numeric_limits<unsigned>::max();
}
} }
else else
{ {
@ -546,6 +542,8 @@ template <class DataFacadeT> class MapMatching final
} }
} }
debug.set_viterbi(model.viterbi, model.pruned);
if (prev_unbroken_timestamps.size() > 0) if (prev_unbroken_timestamps.size() > 0)
{ {
split_points.push_back(prev_unbroken_timestamps.back() + 1); split_points.push_back(prev_unbroken_timestamps.back() + 1);