Walkthrough all Matlab Code Used In This Project

Aus HSHL Mechatronik
Zur Navigation springen Zur Suche springen
Tabelle 17: Storing input data
Name Description
storingcropsteeringandframesoncsvrgb.m Steering values are stored in a CSV file as well as their corresponding frames in RGB.
storingcropsteeringandframesoncsvrgb_calibrated.m Steering values are stored in a CSV file as well as their corresponding calibrated frames in RGB.
Frames_conversion.m

Convert RGB frames to BW, grey, and Canny edge.


Video_conversion.m

Convert RGB videos to BW, grey, and Canny edge.


speedlimit.m

Script which implement minimum and max throttle values

storingcropsteeringandframesoncsvgrayscale.m

Steering values are stored in a CSV file as well as their corresponding frames in gray.

Tabelle 18: Calibration
Name Description
Calibrationchecker.m


Calibrationchecker_video.m


test_camera_calibration.m


256by256frames.m


cali.m


calibration_256_256_script.m


calibration_500by220.m


calibration_500by220_final.m


calibration_with_one_image.m


test_extractedframes_256by256_calibrationpara.m


Bird_eye_view.m


Matlab script to load a parameter and calibrate.


Tabelle 19: Lane Tracking
Name Description
Lane_angle_tracking.m

Matlab script to print right lane angle.

Lane_tracking.m

Matlab script to detect and track lanes.

Lane_tracking_lane_change.m

Matlab script which tracks either the left or right lane and switches the ROI when no lane is detected.

addinglanepliudot.m

Matlab script that displays the steering values as a dot on the screen as well as the detected lanes.

Tabelle 20: python libraries
Name Description
Jtest.py


Jtest0.py

Python script that allows steering by incrementation through buttons 4 and 3.

Jtest01.py

Python script steering without incrementation

Jtest02-object.py

Python script that reads steering_value.txt for autonomous driving and steers automatically when an object is detected.


Jtest4.py

Python script that reads steering_value.txt for autonomous driving.

Tabelle 21: autonomous driving
Name Description
autonomous_driving.m


autonomous_driving_lanetracking.m


RunningtheJetRacerAutonomouslyupdate.m


RunningtheJetRacerAutonomously.m


Tabelle 22: Models
Name Description
Improved_Classified_CNN_Model.mat


Improved_Regression_CNN_Model.mat


Improved_RGB_CNN_model.mat


RGB_CNN_model.mat
RGB_CNN_Regresssion_model.mat


RGB_RGBRect_CNN_Model.mat


RNN_model.mat


trainedNet.mat


Gray_CNN_Classification.mat


Gray_CNN_Regresssion_model.mat


MultiInput_CNN_Model_GrayOnly.mat


CannyEdge_CNN_Regresssion_model.mat


BW_CNN_Regresssion_model.mat


MultiInput_CNN_Model_BW.mat


trainedCNN_Model_1.mat trainedCNN_Model_2.mat trainedCNN_Model_3.mat


trained12vid.fig


Tabelle 23: Training
Name Description
Cleaned_RGB_CNN_Classify_model.m

Matlab script to train the CNN by classification model using RGB frames with more inputs like direction, steering angles etc.

Cleaned_RGB_RGBRect_CNN_Classify_model.m

Matlab script to train the CNN by classification model using RGB rectified frames.

Complete_CNN_Classification.m

Matlab script to train the CNN by classification model using RGB frames.

RGB_CNN_Classify_model.m

Matlab script to train the Classification model using RGB frames.

RGB_CNN_model.m
Matlab script to train the CNN model using RGB frames.

RGB_CNN_Regression_model.m

Matlab script to train the Regression model using RGB frames.

RGB_LSTM_model.m

Matlab script to train the LSTM model using RGB frames.

RGB_Pretrained_model.m

Matlab script to train the Pretrained model using RGB frames.

RGB_Pretrained_model_script.m

Matlab script to train the Pretrained model using RGB frames.

RGB_Reinforcementlearning_model.m

Matlab script to train the reinforcement learning model using RGB frames.

RGB_RNN_model.m

Matlab script to train the RNN using RGB frames.

RGB_Transferlearning_model.m

Matlab script to train the Transferlearning model using RGB frames.

RGBmodel.m

Matlab script to train the model using RGB frames.

Cleaned_Gray_CNN_Classify_model.m

Matlab script to train the Classification model using grey scale frames.

Gray_CNN_model.m

Matlab script to train the CNN model using grey scale frames.

Gray_LSTM_model.m

Matlab script to train the LSTM model using grey scale frames.

Gray_Pretrained_model.m

Matlab script to train the ResNet18 model using grey scale frames.

Gray_Reinforcementlearning_model.m

Matlab script to train the model by Reinforcementlearning using grey scale frames.

Gray_RNN_model.m

Matlab script to train the model by RNN using grey scale frames.

Gray_Transferlearning_model.m

Matlab script to train the model by transfer learning using grey scale frames.

Graymodel.m

Matlab script to train the model using grey scale frames.

greendottraining.m

Matlab script to train the NN with the steering angle represented by a dot on the screen

online_training.m

Matlab script to continue training the model with new data.

online_training_RNN.m

Matlab script to continue training the RNN model with new data.

PretrainedNN.m


resnet_pretrained.m


ResNet18.m

Matlab script to load the pretrained ResNet18.

ResNet18gray.m

Matlab script to train the NN with ResNet18 on grey frames.

ResNet18rgb.m

Matlab script to train the NN with ResNet18 on RGB frames.


Resnet18trainonceperfram.m


traininglaneplusdot.m

Matlab script to train the NN with both the steering angle represented by a dot on the screen and the detected edges.

Tabelle 8: .MAT files
Name Description
cameraParams.mat


cameraCalibrationParams.mat


cameraCalibrationParams0.mat


cameraCalibrationParams_singleangle.mat


calibration256by256.mat


Tabelle 8: zeigeMessdaten
Name Description
Messdaten2MAT.m

Matlab script to convert CSV steering data to a .mat file.

zeigeMessdaten.m

Matlab script to compare actual steering with predicted steering.

zeigeMessdaten_CNN_classify_model.m

Matlab script to compare actual steering with predicted steering for classification model.

zeigeMessdaten_CNN_Regression.m

Matlab script to compare actual steering with predicted steering for regression model.

zeigeMessdaten_grayscale.m

Matlab script to compare actual steering with predicted steering for grayscale model.