jiweibo / ssocr Goto Github PK

View Code? Open in Web Editor NEW

76.0 5.0 45.0 791 KB

Seven Segment Optical Character Recognition

License: GNU General Public License v3.0

Python 100.00%

image-processing python

ssocr's Introduction

SSOCR

Seven Segment Optical Character Recognition

Algorithm

DIGITS_LOOKUP = {
    (1, 1, 1, 1, 1, 1, 0): 0,
    (1, 1, 0, 0, 0, 0, 0): 1,
    (1, 0, 1, 1, 0, 1, 1): 2,
    (1, 1, 1, 0, 0, 1, 1): 3,
    (1, 1, 0, 0, 1, 0, 1): 4,
    (0, 1, 1, 0, 1, 1, 1): 5,
    (0, 1, 1, 1, 1, 1, 1): 6,
    (1, 1, 0, 0, 0, 1, 0): 7,
    (1, 1, 1, 1, 1, 1, 1): 8,
    (1, 1, 1, 0, 1, 1, 1): 9,
    (0, 0, 0, 0, 0, 1, 1): '-'
}

Digital recognition of seven-segment digital tube is relatively simple compared to handwritten numeral.

Detect the existence of the corresponding bit, then encode the image, you can accurately identify the number.

Requirements

opencv
numpy
matplotlib

Setup

git clone https://github.com/jiweibo/SSOCR.git
python ssocr.py images/test1.bmp -s

Results

$ python ssocr.py images\test1.BMP
['-', 3, 0, '.', 3, 7]
$ python ssocr.py images\test2.BMP -s
[1, 7, 7, '.', 7]
$ python ssocr.py images\test3.BMP -s
[0, 7, 8, '.', 3]
$ python ssocr.py images\test4.BMP -s
[0, 7, 2, '.', 6]

Acknowledge

SSOCR

ssocr's People

Contributors

Stargazers

Watchers

Forkers

wonka929 prev-i nanomoretti shovan777 kaizhongkaizhong omarrfarouk awei0905 lrq0701 clarkyeah td0g javism2 asdkj54 jeremydemers shwjdgh34 iamjuush lassoandres bernard9090 tsuyoshiohashi s1ayh0me hexondata-haziq satoshirobatofujimoto stevenchang8 abhishekmishra25 jukha trcclub tuser-code andrewhmtang mfkiwl timemaster7 mhermon iboted nolobot deeyaa01 phateffect hasnul mozg90 huskydj1 szf2020 hyjowd sangtpninjavanvn marcozzxx810 datdoc shawnmolichwier bill0412 dnssoftware

ssocr's Issues

Examples are not working?

Just tried this project with the latest opencv and matplotlib versions and I'm getting a division by zero error?.

Any advise?

Hi

Im having problems on proccessing new images threw your code, i know the problems are with the commented lines in it.
Can you give a help with this?
Thanks a lot

It doesn't work on multi line LCD display

The solution As-Is works for a single line LCD display. How difficult would it be to detect multiline LCD display image and output the results in a list ordered from top down?

Source Images

Can't detect number one

How to Input a test image of seven segment display in code for testing purpose?

Sir,
Greetings!

I wanted to know how to test the code with a test image (.jpg/.png/.bmp) to verify functionality of code for our application.
I tried:

(I am still learning argparse) I included default path - Did not work
parser = argparse.ArgumentParser() parser.add_argument('image_path', help='path to image', default = "C:/Users/Vishw/OneDrive/Desktop/Embtech Internship 2021/Bat_seasoning_rasberryPICV/dep.png")
I tried initializing path below defined load_image() function - Did not work
def load_image(path, show=False): path = "C:/Users/Vishw/OneDrive/Desktop/Embtech Internship 2021/Bat_seasoning_rasberryPICV/dep.png"
gray_img = cv2.imread(path, cv2.IMREAD_GRAYSCALE) h, w = gray_img.shape
blurred = cv2.GaussianBlur(gray_img, (7, 7), 0) if show: cv2.imshow('gray_img', gray_img) cv2.imshow('blurred_img', blurred) return blurred, gray_img

Kindly suggest me what to do.

Thank you!

integer division or modulo by zero

I followed the instruction inputing: python ssocr.py images/test1.bmp -s
Error messages happened:
File "ssocr.py", line 252, in recognize_digits_line_method
if cv2.countNonZero(roi[h - int(3 * width / 4):h, w - int(3 * width / 4):w]) / (9 / 16 * width * width) > 0.65:
ZeroDivisionError: integer division or modulo by zero

I changed the testing images from test2.bmp to test2.bmp, the results were the same.

jiweibo / ssocr Goto Github PK

ssocr's Introduction

SSOCR

Algorithm

Requirements

Setup

Results

Acknowledge

ssocr's People

Contributors

Stargazers

Watchers

Forkers

ssocr's Issues

Source Images

Can't detect number one

Recommend Projects

Recommend Topics

Recommend Org