xtaodada/StarRailCopilot
1
0
Fork 0
mirror of https://github.com/LmeSzinc/StarRailCopilot.git synced 2024-11-15 22:19:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: Convert color space and separate color channels to correctly binarize

Browse Source 
转换为HSV色彩空间，生成橙色和白色掩膜 (遗器信息文字是橙色和纯白色的) 以实现正确的二值化。然后遗器套装就可以正确识别了。
This commit is contained in:
antecanis8 2024-09-28 02:32:29 +08:00
parent bb075e6408
commit 0e6f6c0c06
1 changed files with 30 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
dev_tools/matchTemplate_test.py
View File

@ -10,16 +10,37 @@ class DetectText:
@classmethod @classmethod
def detect_text_areas(cls, image_path): def detect_text_areas(cls, image_path):
imagea = cv2.imread(image_path)
hsv = cv2.cvtColor(imagea, cv2.COLOR_BGR2HSV)
# 定义白色的 HSV 范围，并生成白色掩膜
lower_white = np.array([0, 0, 200])
upper_white = np.array([180, 30, 255])
white_mask = cv2.inRange(hsv, lower_white, upper_white)
# 定义橙色的 HSV 范围，并生成橙色掩膜
lower_orange = np.array([10, 100, 100])
upper_orange = np.array([25, 255, 255])
orange_mask = cv2.inRange(hsv, lower_orange, upper_orange)
# 将两个掩膜进行结合
combined_mask = cv2.bitwise_or(white_mask, orange_mask)
result = cv2.bitwise_not(combined_mask)
# cv2.imshow("result", result)
image = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE) image = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
blurred = cv2.GaussianBlur(image, (9, 9), 0) # blurred = cv2.GaussianBlur(image, (9, 9), 0)
subtracted = cv2.subtract(image, blurred) # subtracted = cv2.subtract(image, blurred)
#
# _, binary = cv2.threshold(
# subtracted, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU
# )
_, binary = cv2.threshold(
subtracted, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU
)
result_inv=cv2.bitwise_not(result)
kernel = np.ones((9, 9), np.uint8) kernel = np.ones((9, 9), np.uint8)
closed = cv2.morphologyEx(binary, cv2.MORPH_CLOSE, kernel) closed = cv2.morphologyEx(result_inv, cv2.MORPH_CLOSE, kernel)
dilated = cv2.dilate(closed, kernel, iterations=2) dilated = cv2.dilate(closed, kernel, iterations=2)
@ -36,8 +57,7 @@ class DetectText:
text_areas.append((x, y, w, h)) text_areas.append((x, y, w, h))
cv2.rectangle(output_image, (x, y), (x + w, y + h), (0, 255, 0), 2) cv2.rectangle(output_image, (x, y), (x + w, y + h), (0, 255, 0), 2)
if cv2.pointPolygonTest(contour, point, measureDist=False) >= 0: if cv2.pointPolygonTest(contour, point, measureDist=False) >= 0:
roi = image[y:y + h, x:x + w] roi = result[y:y + h, x:x + w]
_, binary_roi = cv2.threshold(roi, 150, 255, cv2.THRESH_BINARY_INV)
# cv2.imshow("binary_roi", binary_roi) # cv2.imshow("binary_roi", binary_roi)
# 模板匹配 # 模板匹配
@ -51,8 +71,8 @@ class DetectText:
continue continue
# cv2.imshow("template", template) # cv2.imshow("template", template)
# cv2.imshow("roi", roi) # cv2.imshow("roi", roi)
res = cv2.matchTemplate(binary_roi, template, cv2.TM_CCOEFF_NORMED) res = cv2.matchTemplate(roi, template, cv2.TM_CCOEFF_NORMED)
threshold = 0.8 # 你可以根据需要调整这个阈值 threshold = 0.75 # 你可以根据需要调整这个阈值
loc = np.where(res >= threshold) loc = np.where(res >= threshold)
if loc[0].size > 0: if loc[0].size > 0:
# 如果找到至少一个匹配项，取最大值的索引 # 如果找到至少一个匹配项，取最大值的索引