From c1140803b1a2cd4331020322861afc712ada87d6 Mon Sep 17 00:00:00 2001
From: 283375 <bilibili283375@163.com>
Date: Wed, 30 Aug 2023 20:35:35 +0800
Subject: [PATCH] feat: `FixRects.split_connected`

---
 assets/fix_rects/connected_masked.jpg | Bin 0 -> 1814 bytes
 src/arcaea_offline_ocr/ocr.py         |  51 ++++++++++++++++++++++++++
 2 files changed, 51 insertions(+)
 create mode 100644 assets/fix_rects/connected_masked.jpg

diff --git a/assets/fix_rects/connected_masked.jpg b/assets/fix_rects/connected_masked.jpg
new file mode 100644
index 0000000000000000000000000000000000000000..8e08242d3009d0da8e29a7e02791e14a1728db17
GIT binary patch
literal 1814
zcmX9;4LqCI8b9%o2r^$S%5GC16{cvZW>b@uqP5$(SbdhXlnjeZs}(DGrzpy0s8D{F
zektun)K_X%X$Y;cWv$jnY!Z5Jt%ycxyqY(8_u75V?>y(6-+7+@|2gM*&OxV8888kC
zI2r(OH~?^11LzF!16sH(+0t5DqNBZ~b#-*Kb?~}){NJ5G)WZ{q1Uz2PK#!=uCD_c+
zK;Lj{fj$BTdVm64aX1rziU1h{*Zmt4WB9+|w6t|}@dV7*05j+|#yGXLF(zG%2D8q=
z_JNLxuBpwwLwK{0Qv}=eZSJ`h*Y)gt?+LeuK2_Lz#GkoH)Hk=V++nrbf#T@&(SFYZ
z)Q>;$@;mGw5P0NhP}rBp!y_W2q7zOhCNaKBPC1)#?tCUY>)X8if=h)(#g{9qu2get
zYOmJaxOuDTc5};}yB(cf-97yKy*~{MiiU<CKN<P?+4%Dp6O&WZlDYYp|NY;qg+E^}
zuD)Gcf2Z7lH{YWYptp5xG@yZ&{02B?s}bed$8WPO9`)~d?%B$Ues0tLw2SF5vXaJI
zxt;!DxMIlCZBN^BOUt!R4}fzJFa)w5nj#R&qbq`F@UEHxvi>ZaCVVd#cIK6tGv<v%
za!R1>(!VMff+mtKSf(+?=7&v4KYnht8S?YQjk#-OQ76J7#JFBV98`Uv5z`gH0!v89
zhGci+01EQ?LND@+i#JXdBiEmaXn&Z0!l91UF!rGJ#n}7&%E&_`$sp;$;YoK>p6*aD
z3QWF7!Z5FL&2|QIFieI5#A5-mL%|eX(c<%zCV@Hxo&BRqqn<42l<EKqHWSoWJC{-L
zr|PU~aoq@Z;v>c?StGKYu1tp$QJ@=yf{O;(coh}d^N*eed-!-l>TUL!0mqXuC0VRW
z>-r*><u>LDlXOKfbJeCdDW<4=bk7DY-&6G|3Um&rD-j|#wy3$@$524Pp+M-1c&JN!
zYUVcM)wO+;AK=TeJ`19a4K6i(vyo9ow5KtPMdlrQE5iSH!-~^3Jip=Bj)gGBnQYtv
zv4UaGB_&7Q=0lr(y~gGZH+JQ`8@~3_vG01id&-=$V_7Vw;mJzdOS6`zLYLpV4_gIR
zj2`41fk&0^)IV}|afWBjFfb1MV4g40n6Rxyrl}lrMU8TMJ`^vgcoX|d75bU$t<e;V
z!mlek$`9tgb&!&Ii&TwXH)W6pQztcgzs#Ya2vU3*ymOIE+717(>bnrhv<_EP%ZDaP
z&Sk|{NRrq5<7?cD<HSd(%vXVCzT@>-A-juKkPwU+4+$OLmBSy2<wMOdMKsYLG3ch1
zQ!gTKb4zn^Go*42&g5of)J-*4oxM<O=bKbqY1!6X`eZ^AvrbmTiD73k;_dgci>!r$
zq<^QX?39UctTN}|95t`&AwgyPE7U1zeUK6wt&M`%_^$BCesNOBVdJANPpYIS__};E
zjHUbn1)bkL7p(J!k8f(zJEe*M2==O)V<2XH<rgqZWr(Dz%li)9WhV$xFhiDbK2`8b
zb~R||-BM|a$!7Ekg`ahNOzj+VvZR3IIHF}Y1qnqcP++*_0zR*RL&lOEA~;Ex1kM-A
z4=JSc`0MV;aXQb|ms%7194GbUuU>~o=iF_ozT8xLK6BqO!_1J_m_X%KS4v$l<#!*a
zY=XKFVxR!FZQ4Lg`5Iy&R)(eO+WjanW-p0MVzXmbQ6O?F=St2leKV?0@|XJ;U=f^N
z(~KlCl%YK0y|tIsP27Nn;4vfk8@f8_rV@KeqB0b$SgC31`;e@X7LS6wbOFO{DpO;<
zp@e=z!JIXm%bbVc=$*7xv65PFj>n%_;!7FhtnA+YNfg+452(WA#xdTicrlk|Nn^5L
zyGi-9N+)~II10WBE^ELk)Z#%GNMw>6_W@inNBc1O+n(xqx({)M>AF1haPWj@_fU6V
z)sTJr<15bh?e<1*UKHGs!X7M*c_ZRQ3q?Ue0Sbr&Wgy~8VrY!%@<!USbeev891*u?
z3EZXyU9>K07M&vT8E>4;sz&aG&~Kuki!WbNQL;b7PLxbz@L<ZDe2#MtZ~jeg|K7MI
zj?-3oBT#&M<$|cKD(8;Oyd&|Pfk)3LQHB3`vPovOKtdxF$nKLn34_w<ylAXS{jm(j
zdRsIv(L4LQo24XeI#WND`<o>B)<f>Ei_Z2`Q_;ph$7lX{M0(LQ?qry-ZzS=xMjvy;
zN?h3LV+BVUA=?Z@kRpU$y3-ckk-VxMtjIX<o||{UNP-;QW|k5WZgGrxBap2epQ=A;
zXcAM~*Sne_-k(aSg@lm;nM`~svk~8CfPEbr4Av$J9+W*EfG*YLm-TxnW_cH`rFe}V
iPP<;_%<BA|_)F<PQXe<wSz240Ltx#<rwb$yI{g=2oPYZO

literal 0
HcmV?d00001

diff --git a/src/arcaea_offline_ocr/ocr.py b/src/arcaea_offline_ocr/ocr.py
index abda786..b7b6598 100644
--- a/src/arcaea_offline_ocr/ocr.py
+++ b/src/arcaea_offline_ocr/ocr.py
@@ -70,6 +70,57 @@ class FixRects:
         return_rects.extend(new_rects)
         return return_rects
 
+    @staticmethod
+    def split_connected(
+        img_masked: Mat,
+        rects: Sequence[Tuple[int, int, int, int]],
+        rect_wh_ratio: float = 1.05,
+        width_range_ratio: float = 0.1,
+    ):
+        connected_rects = []
+        new_rects = []
+        for rect in rects:
+            rx, ry, rw, rh = rect
+            if rw / rh > rect_wh_ratio:
+                # consider this is a connected contour
+                connected_rects.append(rect)
+
+                # find the thinnest part
+                border_ignore = round(rw * width_range_ratio)
+                img_cropped = crop_xywh(
+                    img_masked,
+                    (border_ignore, ry, rw - border_ignore, rh),
+                )
+                white_pixels = {}  # dict[x, white_pixel_number]
+                for i in range(img_cropped.shape[1]):
+                    col = img_cropped[:, i]
+                    white_pixels[rx + border_ignore + i] = np.count_nonzero(col > 200)
+                least_white_pixels = min(list(white_pixels.values()))
+                x_values = [
+                    x
+                    for x, pixel in white_pixels.items()
+                    if pixel == least_white_pixels
+                ]
+                # select only middle values
+                x_mean = np.mean(x_values)
+                x_std = np.std(x_values)
+                x_values = [
+                    x
+                    for x in x_values
+                    if x_mean - x_std * 1.5 <= x <= x_mean + x_std * 1.5
+                ]
+                x_mid = round(np.median(x_values))
+
+                # split the rect
+                new_rects.extend(
+                    [(rx, ry, x_mid, rh), (rx + x_mid, ry, rw - x_mid, rh)]
+                )
+
+        return_rects = deepcopy(rects)
+        return_rects = [r for r in rects if r not in connected_rects]
+        return_rects.extend(new_rects)
+        return return_rects
+
 
 def resize_fill_square(img: Mat, target: int = 20):
     h, w = img.shape[:2]