Improving the robustness of all three algorithms.

Samo Penic

2018-11-17 d5c694ac711ca3b434bf16bd920b90d1a7e758c4

Improving the robustness of all three algorithms.

 Ocr.py

@@ -222,7 +222,7 @@
            sid_mask=self.settings.get("sid_mask", None)
        es,err,warn = getSID(
            self.img[
                int(0.045 * self.imgHeight) : int(0.085 * self.imgHeight),
                int(0.04 * self.imgHeight) : int(0.095 * self.imgHeight),
                int(0.7 * self.imgWidth) : int(0.99 * self.imgWidth),
            ],
            self.sid_classifier,

 aoiOcr.py

@@ -3,14 +3,24 @@

from glob import glob

settings = {"sid_mask": "61xx0xxx", "answer_treshold": 0.25}
settings = {"sid_mask": "64xx0xxx", "answer_treshold": 0.25}
classifier = joblib.load("filename.joblib")

#p = Paper(filename="testpage300dpi_scan1.png")
#p=Paper(filename='sizif111.tif', sid_classifier=classifier, settings=settings)
# p = Paper(filename="testpage300dpi_scan1.png")
#p=Paper(filename='sizif111.tif', sid_classifier=classifier, settings={"sid_mask": "11xx0xxx", "answer_treshold": 0.25})
#p=Paper(filename='processed_scans/20141016095134535_0006.tif', sid_classifier=classifier, settings=settings)
#p=Paper(filename='processed_scans/20151111080408825_0001.tif', sid_classifier=classifier, settings=settings)
p=Paper(filename='processed_scans/20151028145444607_0028.tif', sid_classifier=classifier, settings=settings)
#p = Paper(filename="processed_scans/20151111080408825_0001.tif",sid_classifier=classifier,settings=settings,)
#p=Paper(filename='processed_scans/20151028145444607_0028.tif', sid_classifier=classifier, settings=settings)
pa = [
    "processed_scans/20141016095134535_0006.tif",
    "processed_scans/20141016095134535_0028.tif",
    "processed_scans/20141016095134535_0028.tif",
    "processed_scans/20141016095134535_0037.tif",
    "processed_scans/20141021095744144_0005.tif",
    "processed_scans/20141021095744144_0009.tif",
    "processed_scans/20141028095553745_0018.tif",
]
p=Paper(filename=pa[6], sid_classifier=classifier, settings=settings)

# print(p.QRData)
# print(p.errors)
@@ -23,8 +33,13 @@


print(p.get_paper_ocr_data())
exit(0)
filelist = glob("processed_scans/*.tif")
for f in filelist:
    print(f,Paper(filename=f, sid_classifier=classifier, settings=settings).get_paper_ocr_data())

filelist = glob("processed_scans/*.tif")
for f in sorted(filelist):
    print("processing: {}".format(f))
    print(
        f,
        Paper(
            filename=f, sid_classifier=classifier, settings=settings
        ).get_paper_ocr_data(),
    )

 sid_process.py

@@ -57,6 +57,13 @@
def kernel(x, y):
    return np.ones((x, y), np.uint8)

def sid_compare(sid_no, sid_mask):
    for s,es in zip(sid_mask,sid_no):
        if s!='x' and s!=es:
            return False
    return True



def segment_by_contours(image, sorted_ctrs, classifier):
    sid_no = ""
@@ -77,7 +84,7 @@
    return sid_no


def segment_by_sid_len(image, sid_mask, classifier):
def segment_by_sid_len(image, original_image, sid_mask, classifier):
    sid_no = ""
    sid_len = len(sid_mask)
    if sid_mask[0] == "1":
@@ -86,9 +93,11 @@
        move_left = 55
    else:
        move_left = 0
        # Remove noise
    image2 = cv2.morphologyEx(original_image, cv2.MORPH_OPEN, kernel(2, 2), iterations=7)
    # find biggest block of pixels

    image1 = cv2.morphologyEx(image, cv2.MORPH_DILATE, kernel(5, 25), iterations=4)
    image1 = cv2.morphologyEx(image2, cv2.MORPH_DILATE, kernel(5, 25), iterations=4)
    image1=img_as_ubyte(image1>50)
    cv2.imwrite("sidblock1.png", image1)
    im2, ctrs, hier = cv2.findContours(
        image1.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE
@@ -97,7 +106,7 @@
        ctrs, key=lambda ctr: cv2.contourArea(ctr)
    )  # get bigges contour
    x, y, w, h = cv2.boundingRect(sorted_ctrs[-1])
    image = image[y : y + h, x + 25 - move_left : x + w - 25]
    image = image[y : y + h, x + 25 - move_left : x + w - 40] #+25,-25
    cv2.imwrite("sidblock2.png", image)
    imgHeight, imgWidth = image.shape[0:2]
    numWidth = int(imgWidth / (sid_len))
@@ -111,14 +120,60 @@
        sid_no = sid_no + str(classifier.predict(num.reshape(1, -1) / 255.0)[0])
    return sid_no

def segment_by_7segments(image,original_image,sid_mask,classifier):
    block_image = cv2.morphologyEx(original_image, cv2.MORPH_CLOSE, kernel(2, 2), iterations=10)
    block_image =img_as_ubyte(block_image<50)
    cv2.imwrite("sid_3rd1.png", block_image)
    template = cv2.imread("template-8.png", 0)
    w, h = template.shape[::-1]
    res = cv2.matchTemplate(block_image, template, cv2.TM_CCOEFF_NORMED)
    loc = np.where(res >= 0.75)
    cimg = cv2.cvtColor(image, cv2.COLOR_GRAY2BGR)
    loc_filtered_x=[]
    loc_filtered_y=[]
    for pt in zip(*loc[::-1]):
        pt=(pt[0]-10,pt[1]-10)
        loc_filtered_y.append(pt[1])
        loc_filtered_x.append(pt[0])
#        points.append(pt)
    #filter points
    if(len(loc_filtered_x)==0):
        return ""
    loc_filtered_x, loc_filtered_y = zip(
                *sorted(zip(loc_filtered_x, loc_filtered_y))
            )
    a = np.diff(loc_filtered_x) > int(w/2)
    a = np.append(a, True)
    loc_filtered_x = np.array(loc_filtered_x)
    loc_filtered_y = np.array(loc_filtered_y)
    points = [loc_filtered_y[a], loc_filtered_x[a]]
    for pt in zip(*points[::-1]):
        cv2.rectangle(cimg, pt, (pt[0] + w, pt[1] + h), (0, 255, 255), 2)
    cv2.imwrite("sid_3rd2.png", cimg)

    sid_no=""
    for i,pt in enumerate(zip(*points[::-1])):
        num=image[pt[1]:pt[1] + h, pt[0]:pt[0]+w]
        #cv2.imwrite("sid_3no_{}.png".format(i), num)
        num = img_as_ubyte(num < 128)
        try:
            num = cv2.resize(num, (32, 32))
        except:
            return ""
        cv2.imwrite("sid_3no_{}.png".format(i), num)
        sid_no = sid_no + str(classifier.predict(num.reshape(1, -1) / 255.0)[0])

    return sid_no

def getSID(image, classifier, sid_mask):
    sid_warn = []
    sid_err=[]
    image = 255 - image
    image_original=image.copy()
    image = img_as_ubyte(image > 100)
    cv2.imwrite("enSID0.png", image)
    # Remove noise
    image = cv2.morphologyEx(image, cv2.MORPH_OPEN, kernel(2, 2), iterations=1)
    image = cv2.morphologyEx(image, cv2.MORPH_OPEN, kernel(2, 2), iterations=3)
    # Closing. Connect non connected parts
    image = cv2.morphologyEx(image, cv2.MORPH_CLOSE, kernel(5, 3), iterations=4)
    # Again noise removal after closing
@@ -144,14 +199,21 @@
    )
    sorted_ctrs = sorted(ctrs, key=lambda ctr: cv2.boundingRect(ctr)[0])

    sid_no = ""
    print(len(sid_mask), len(sorted_ctrs))
    sid_no = segment_by_contours(
        image, sorted_ctrs[1:], classifier
    )  # we remove largest contour that surrounds whole image
    print(sid_no)
    if len(sid_no) != len(sid_mask):
        #print("Ooops have to find another way")
    if len(sid_no) != len(sid_mask) or not sid_compare(sid_no,sid_mask):
        sid_warn.append("Trying second SID algorithm.")
        sid_no = segment_by_sid_len(image, sid_mask, classifier)
    return (sid_no, [], sid_warn)
        sid_no = segment_by_7segments(image, image_original, sid_mask, classifier)
    print(sid_no)
    if(len(sid_no))!=len(sid_mask):
        sid_no = segment_by_sid_len(image, image_original, sid_mask, classifier)
        sid_warn.append("Trying third SID algorithm.")


    if not sid_compare(sid_no, sid_mask):
            sid_err=['Wrong SID!']

    return (sid_no, sid_err, sid_warn)

 template-8.png