PyDet

Module with supplier functions oriented around computer vision

To use this module, download this repository and add it to the folder you need PyDet for. The necessary files/folders are:

• build
• models
• pybind11
• yolov5-master
• CMakeLists.txt
• Makefile
• cmake_install.cmake
• filters.cpp
• pydet.py

To download, run git clone https://github.com/v2pir/PyDet.git

After adding this to your folder, to use PyDet, simply type:

import pydet as pd

You can then use the functions defined.

Documentation

Detect

• detectBlob(): detects blob in image
• detectTumor(): Uses a pre-trained YOLOv5 model to detect brain tumors in MRI scans
• detectPet(): Uses a pre-trained YOLOv5 model to detect animals in images
• detectHuman(): Uses a pre-trained YOLOv5 model to detect humans in images
• detectText(): Uses OCR to detect text in images

Colors

• findSpecificColor(): given an HSV range and an image parameter (string to its relative path), it will detect any contour within that HSV range
• findColor(): pre-defined HSV ranges for common colors including: ORANGE, RED, GREEN, BLUE, PINK, YELLOW

Random set of functions

• splitVideo(): splits a video into frames
• getOrientation(): returns the yaw, pitch, roll, and camera position, of a given object in the image (3d array of object dimensions + 2d array of coordinates of the object in the image is required)
• LinRegLossFunc(): returns the losses and weights of a linear regression model

Filters

• saturate(): will saturate an image by a given intensity (the function defined in filters.cpp which is why pybind11/CMake is needed)