The Cattle Counter Application invloves feeding a image or a video of cattle and the application will detect the cattle present in the image/video. The application afte processing the image will return the count of the detected cattle.
- OpenCV
- TensorFlow
Collected images from various online sources and merged them into a single dataset. The dataset consists of two folders. The training dataset and the validation dataset. The training dataset consists of 50 images of various cattle 50 images of humans to represent non-cattle along with the dataset having 25 images of various cattle 50 images of humans.
Using a seqential model to train the data. The below presented flowchart represents the the model training process.
graph TD;
Dataset-->Sequential_Model;
Sequential_Model-->Save_best_model;
After training the model, the accuracy of the dataset was realised to be. Along with validation accuracy to be.
After the best model has been saved, a new image is fed into the model where the model detects the cattle present in the new image. Using OpenCV, we draw bounding boxes around the detected cattle. After call the cattle have been detected, a counter is placed to keep a count on the number of bounding boxes present. The prpcess can be defined by the below presented flowchart.
graph TD;
Image-->Predict_through_saved_model;
Predict_through_saved_model-->Detect_cattle;
Detect_cattle-->Draw_box_around_detected_cattle;
Draw_box_around_detected_cattle-->Add_counter_according_to_number_of_boxes;
Add_counter_according_to_number_of_boxes-->Print_Counter;
The results of the trained model has been provided in the uploaded colab notebook.