LIDAR Filters This project “LIDAR Filters” is designed to reduce noise in LIDAR scans. There are two types of filters:

1. Range Filter - This filter crops extreme values so that all values lie within an acceptable range of values
2. Median Filter - This filter smoothens the scan values by taking the median of 'D' previous scan values

PREREQUISITES:

1. This project requires Python 2.7.x with numpy module.

SETUP: Step 1: To utilize these filters, just copy the source files into your project and import them: from import RangeFilter from import MedianFilter Step 2: The error messages in these classes are externalized into errormessages.py. Also copy and import this class: from import ErrorMessages

USING THE RANGE FILTER: Step 1. Initialize the RangeFilter object. This object gives you the option to modify the default range minimum (0.03) and range maximum (50) values. Ex: rangefilter = RangeFilter(range_min = 1, range_max = 10) rangefilter = RangeFilter() Step 2. Pass valid scan lists one by one to the update method of this object. This method returns a list of range cropped scan values. Ex. rangefilter.update([1.5, 2.0, 2.5, 3.0])

USING THE MEDIAN FILTER: Step 1. Initialize the MedianFilter object by passing a valid D value (D must be a positive integer). This object gives you the option to add historic scans that you may want to be considered in the smoothening future scans. This data must be a list of scans of the same length. Based on the given D value, this historic scans will be used to calculate the median of future scans. Ex: medianfilter = MedianFilter(2, [[1, 2 , 3], [3, 2, 5]] medianfilter = MedianFilter(2) Step 2. Pass valid scan lists one by one (of equal lengths) to the update method of this object. This method returns a list of smoothend values Ex. medianfilter.update([5,2,6]) RUNNING TEST CASES: The unit test files for the source files can be found under “tests” directory. These tests use a utility ExtendedTestCase class to check if the raised error message matches the expected error message.

CONSTRAINTS:

1. A valid scan is a list of Integer/Float values.
2. All scans sent to an instance of MedianFilter object must be of equal length.
3. Historic scan data sent to initialize the MedianFilter object must be a list of valid scans of equal length.
4. D value sent to MedianFilter must be a positive integer.
5. Custom range minimum and maximum values must be Integer/Float values.

REFERENCES:

1. ExtendedTestCase class has been taken from a blog in StackOverFlow.
2. Referred https://docs.python-guide.org/writing/structure/ to structure my project.
3. Referred https://docs.python.org/2/library/unittest.html to understand the Unit Testing framework in python.