Idea which I want to impliment for the task Line detection For line detection I used HoughLinesP opencv library.i am getting good result in HoughLinesP as compare to HoughLines . Text detection For text detection the most accurate was pytesseract library from python.although the text is not visible in last 3 or 4 images but I got good result using pytesseract. Object detection Object detecton I used YOLO V5 model the data is very less for train but I did data agmentation and improve the accuracy littlebit .
HoughLinesP code
Output
Pytesseract Code
Output
labeling ,Data augmentation,YOLO train Train and validation I split the data set into train and validation.8 image for train folder and 4 for validation Labeling For labeling I used a website www.makesense.ai and labeling all the train images and val images Data augmentation Data is very less for YOLOV5 model so I did some data augmention method like rotated the image clockwise and labeling all the rotated images Upload data and modified Yolo file I uploaded the zip dataset into the collab and modified the COCO128.yaml Training YOLO V5 model Train the model I set the batch size is 60 and epoch is 1000.