scieee AI-readable full text Open interactive document viewer

Obstacle Detection System Dataset

Dini, Fabrizio

Abstract

This dataset provides data from the Obstacle Detection System installed on Tram 1013 operating in service in Florence tramway system. It is equipped with a set of sensors (RADAR, LiDAR, Cameras) and collects datasets from relevant measurements. The system processes the sensor data and determines the cinematic information of detected objects. Raw data containing measurements of targets detected in a specified section of the tramway corridor and for different time periods is provided.

Full text

1 README ODT Overview Tram 1013 operating in revenue service in Florence tramway system is equipped with a set of sensors (RADAR, LiDAR, Cameras) and collects datasets from relevant measurements. The first and second dataset associated to ODT system is made of the all sensors’ outputs. • RADAR: used in cluster mode • LiDAR: to collect a dots cloud • Cameras: to collect videos. This dataset contains sensitive information and is not included in this dataset. Though first and second dataset associated to ODT refer to different measurement sessions and are in the following data format: • .csv files ODT system processes data from sensor with many complex algorithms which detect and track position and trajectory of many objects standing still or moving in front of tram 1013. Dataset Contents/Dataset Description The files that compose the ODT dataset are: ThalesARS_CompleteObjectsFrame.csv ThalesIMU_GetDataFE_response.csv ThalesQM8_All3Returns_Packet.csv The following tables report the fields and meaning of the content of the above files. 2 Table 1 ThalesARS_CompleteObjectsFrame.csv file structure (ODT RADAR) Item Type Description numberOfObjects int64 Number of objects detected by the RADAR. measurementCycleCounter int64 Number of cycles or pulses that have been emitted and received, indicating the sequencing of measurement cycles in a radar operation. interfaceVersion uint8 Version of the interface. objectListReceptionTime_ns int64 Timestamp in nanoseconds at which a list of detected objects objectFullInfos ARS_ ObjectFullInformation[ ] Refers to the detailed set of information about each detected cluster, including position, velocity, size, and other attributes, which are crucial for understanding and analyzing the RADAR's observations of objects in its vicinity. id uint8 ID of the acquired object longitudinalDistance_m float64 Longitudinal distance of the acquired object expressed in meters. lateralDistance_m float64 Lateral distance of the acquired cluster expressed in meters. dynamicProperty uint8 Dynamic property of the message: stationary cluster, moving cluster, etc.. radarCrossSection_dBm2 float32 Radar cross section of the acquired cluster. longitudinalDistance_ rms float32 Longitudinal distance of the acquired cluster expressed in rms. lateralDistance_ rms float32 Lateral distance of the acquired cluster expressed in rms. longitudinalRelativeVelocity_ rms float32 rms of longitudinal relative velocity lateralRelativeVelocity_ rms float32 rms of lateral relative velocity longitudinalRelativeAcceleration_ rms float32 rms of longitudinal relative acceleration lateralRelativeAcceleration_ rms float32 rms of lateral relative acceleration 3 orientationAngle_ rms float32 rms of orientation angle probabilityOfExistance uint8 0x00: invalid; 0x01: <25%; 0x02: <50%; 0x03: <75%; 0x04: <90%; 0x05: <99%; 0x06: <99.9%; 0x07: <=100% measurementState uint8 0x0: deleted; 0x1: new; 0x2: measured; 0x3: predicted; 0x4: deleted for merge; 0x5: new from merge longitudinalRelativeAcceleration_m_s2 float32 Longitudinal relative acceleration ms^-2 lateralRelativeAcceleration_m_s2 float32 Lateral relative acceleration in ms^- 2 objectClass uint8 0x0: point; 0x1: car; 0x2: truck; 0x3: unused; 0x4: motorcycle; 0x5: bicycle; 0x6: wide; 0x7: reserved orientationAngle_deg float32 orientation angle objectLength_m float32 object length objectWidth_m float32 object width Frequency is 12.5 Hz. 4 Table 2 ThalesQM8_All3Returns_Packet.csv dataset (ODT LiDAR) Item Type Description timestamp_seconds uint32 timestamp in seconds timestamp_nanoseconds uint32 fractional part of the timestamp, in nanoseconds firingData QM8_All3Returns_FiringData[50] A vector of 50 firing data, each specified as follows: position uint16 a number in [0; 10399] encoding the radial position of the sampled points. distancesIntensities QM8_LaserDistancesIntensities[3] A vector of 3 data structure, one for each return, specified as follows: distances uint32[8] a vector of 8 distances, one for each layer intensities uint8[8] a vector of 8 intensities, one for each layer Table 3 - ThalesIMU_GetDataFE_response.csv dataset (ODT IMU) Item Type Description rosbagTimestamp uint64 Epoch timestamp in nanoseconds. epoch_ms float64 Epoch timestamp in ms. packetCounter uint64 Packet counter used to count the received IMU messages. quaternionW float64 QuaternionW component that represents the cosine of half the rotation angle, indicating the magnitude of the rotation about a specific axis in 3D space. quaternionX float64 QuaternionX component that represents the cosine of half the rotation angle, indicating the magnitude of the rotation about a specific axis in 3D space. quaternionY float64 QuaternionY component that represents the cosine of half the rotation angle, indicating the m agnitude of the rotation about a specific axis in 3D space. quaternionZ float64 QuaternionZ component that represents the cosine of half the rotation angle, indicating the magnitude of the rotation about a specific axis in 3D space. euler_yaw_angle_deg float64 Euler yaw angle of an IMU, expressed in degrees: defines the object's rotation around its vertical axis, representing its heading or direction relative to a reference direction like north. 5 euler_pitch_angle_deg float64 Euler pitch angle of an IMU, expressed in degrees, measures the object's tilt forward or backward around its lateral axis, indicating its inclination relative the horizontal plane. euler_roll_angle_deg float64 Euler roll angle of an IMU, expressed in degrees, describes the object's rotation around its longitudinal axis, indicating its sideways tilt relative to the horizontal plane. accelerometerX_g float64 Acceleration along X axis expressed in g. accelerometerY_g float64 Acceleration along Y axis expressed in g. accelerometerZ_g float64 Acceleration along Z axis expressed in g. magnetometerX_Gauss float64 Measures the magnetic field's strength along the device's X-axis in Gauss, helping to determine orientation relative to Earth's magnetic field. magnetometerY_Gauss float64 Measures the magnetic field's strength along the device's Y-axis in Gauss, helping to determine orientation relative to Earth's magnetic field. magnetometerZ_Gauss float64 Measures the magnetic field's strength along the device's Z-axis in Gauss, helping to determine orientation relative to Earth's magnetic field. gyroscopeX_dps float64 Measures the rate of rotation around the device's X-axis in degrees per second, tracking angular velocity. gyroscopeY_dps float64 Measures the rate of rotation around the device's Y-axis in degrees per second, tracking angular velocity. gyroscopeZ_dps float64 Measures the rate of rotation around the device's Z-axis in degrees per second, tracking angular velocity. temperature_C float64 Temperature in °C detected by the sensor. The following is a geometrical info about the LiDAR measurement system 6 Getting Started The dataset is provided in .csv format, so no special software or prerequisites are required to access or use it. Download & Usage • git clone https://github.com/VERGE-PROJECT/GTSI-HITACHI-ODT-dataset The files are zipped, divided per folder and, because github limitation (file size less than 25MB), splitted in several files. Citation NA