1 # Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
3 @page dwx_roadmark_classification_sample Road Mark Classification Sample
6 @section dwx_roadmark_classification_description Description
8 The road mark classification sample demonstrates how to use the NVIDIA<sup>®</sup>
9 proprietary deep neural network (DNN) RoadMarkNet to perform classification on detected road markings from a given frame <br>. It classifies the type of road markings that include different
10 types of arrow marks, speed limit mark and its associated value, HOV mark, bike mark, hump mark
11 and a variety of text marks.
13 A user assigned threshold value sets the minimum confidence for the road mark classification to be considered valid.
15 @subsection dwx_roadmark_classification_details Sensor Details
17 The image datasets used to train RoadMarkNet have been captured by a View Sekonix Camera Module (SF3324/5)
18 with AR0231 RCCB sensor.<br>
19 The camera is mounted high up at the rear view mirror position. Demo videos are captured at 2.3 MP and down-sampled to 960 x 604.
21 To achieve the best road mark classification performance, adopt a similar camera setup and align
22 the video center vertically with the horizon before recording new videos.
24 @section dwx_roadmark_classification_running Running the Sample
26 The road mark classification sample, `sample_roadmark_classification` accepts the following optional parameters.<br>
27 If none are specified, it will perform classification on supplied pre-recorded video. **Note:** Do not include angle brackets < and > in the parameters below.
29 ./sample_roadmark_classification --camera-type=<camera>
30 --camera-group=<a|b|c|d>
31 --camera-index=<0|1|2|3>
32 --input-type=<video|camera>
34 --classificationThreshold=<floating-point number in (0, 1)>
35 --video=<path/to/video>
39 --camera-type=<camera>
40 Is a supported AR0231 `RCCB` sensor.
41 Only applicable if '--input-type=camera'.
42 Default value: ar0231-rccb-bae-sf3324
44 --camera-group=<a|b|c|d>
45 Is the group where the camera is connected to.
46 Only applicable if '--input-type=camera'.
49 --camera-index=<0|1|2|3>
50 Indicates the camera index on the given port.
53 --input-type=<video|camera>
54 Defines if the input is from live camera or from a recorded video.
55 Live camera is only supported on On NVIDIA DRIVE platform.
59 Defines a Region of Interest (ROI) where landmark detections occur:
64 Default value: No ROI.
66 --classificationThreshold=<floating-point number in (0, 1)>
67 The classification threshold parameter is used to determine the validity of the road mark
68 type generated by the network. If there is no type above the threshold parameter, then the
69 road mark is assigned as background.
72 --video=<path/to/video>
73 Is the absolute or relative path of a raw/h264/mp4 recording.
74 Default value: path/to/data/samples/recordings/suburb0/video_0_roof_front_120.mp4.
76 @subsection dwx_roadmark_classification_examples Examples
78 #### To run the sample on Linux
80 ./sample_roadmark_classification --video=<video file.h264>
83 ./sample_roadmark_classification --video=<video file.raw>
86 ./sample_roadmark_classification --video=<video file.mp4>
88 #### To run the sample on an NVIDIA DRIVE platform with cameras
90 ./sample_roadmark_classification --input-type=camera --camera-type=<camera_type> --camera-group=<camera_group>
92 where `<camera type>` is a supported `RCCB` sensor.
93 See @ref supported_sensors for the list of supported cameras for each platform.
95 @section dwx_roadmark_classification_sample_output Output
97 The sample app creates a window and overlays the bounding box of road mark detection from landmark detector. The type of road marking and it's associated confidence from roadmark classifier is displayed on top of each bounding box.<br>
99 
101 @section dwx_roadmark_classification_more Additional Information
103 For more details see @ref landmarks_mainsection.