Welcome to IMX 415 Camera Module repository
6. Color Recognition
7. Color Recognition
1. Introduction to the HSV color space:
The RGB color space is the color space everyone is most familiar with, that is, the three primary color space. Any color can be formed by mixing these three colors. However, generally the effective processing of images in a color space is carried out in the HSV space. HSV (Hue, Saturation, Value) is a color space created based on the intuitive characteristics of color, also known as the hexcone model.
Why choose the HSV space instead of the RGB space? For images, identifying the corresponding color is feasible in the RGB space, the HSV space, or other color spaces. The reason HSV is chosen is that the hue represented by H can basically determine a certain color, and then combined with the saturation and value information, it is judged to be greater than a certain threshold. RGB, on the other hand, is composed of three components and requires judging the contribution ratio of each component. That is, the recognition range of the HSV space is wider and more convenient.
HSV color space model
2. Conversion of three color spaces (gray BGR HSV):
There are more than 150 methods for color space conversion in OpenCV, but there are only two that we commonly use, namely BGR->Gray and BGR->HSV. Note that Gray and HSV cannot be converted to each other. Color space conversion: cv2.cvtColor(input_image, flag)
BGR->Gray: the flag is cv2.COLOR_BGR2GRAY
BGR->HSV: the flag is cv2.COLOR_BGR2HSV
The value range of the HSV color space in OpenCV:
H [0, 179] S [0, 255] V [0, 255]
Value ranges of common colors
According to the above ranges, our goal is to identify whether it is our set color range, and first replace this color with white, replace non-this-color with black, and finally replace the white part with the original image.
Code path: /home/pi/Yahboom_Project/1.OpenCV course/04 advanced tutorial/Color_recognition_USB.ipynb
xxxxxxxxxx#bgr8 to jpeg formatimport enumimport cv2def bgr8_to_jpeg(value, quality=75):return bytes(cv2.imencode('.jpg', value)[1])#Camera component displayimport traitletsimport ipywidgets.widgets as widgetsimport time# Thread function operation libraryimport threadingimport inspectimport ctypesorigin_widget = widgets.Image(format='jpeg', width=320, height=240)mask_widget = widgets.Image(format='jpeg',width=320, height=240)result_widget = widgets.Image(format='jpeg',width=320, height=240)# create a horizontal box container to place the image widget next to eachotherimage_container = widgets.HBox([origin_widget, mask_widget, result_widget])display(image_container)#Thread related functionsdef _async_raise(tid, exctype):"""raises the exception, performs cleanup if needed"""tid = ctypes.c_long(tid)if not inspect.isclass(exctype):exctype = type(exctype)res = ctypes.pythonapi.PyThreadState_SetAsyncExc(tid, ctypes.py_object(exctype))if res == 0:raise ValueError("invalid thread id")elif res != 1:# """if it returns a number greater than one, you're in trouble,# and you should call it again with exc=NULL to revert the effect"""ctypes.pythonapi.PyThreadState_SetAsyncExc(tid, None)def stop_thread(thread):_async_raise(thread.ident, SystemExit)#Main process functionimport cv2import numpy as npimport ipywidgets.widgets as widgetscap = cv2.VideoCapture(0)cap.set(3, 640)cap.set(4, 480)cap.set(5, 120) #Set frame ratecap.set(cv2.CAP_PROP_FOURCC, cv2.VideoWriter.fourcc('M', 'J', 'P', 'G'))# image.set(cv2.CAP_PROP_BRIGHTNESS, 40) #Set brightness -64 - 64 0.0# image.set(cv2.CAP_PROP_CONTRAST, 50) #Set contrast -64 - 64 2.0# image.set(cv2.CAP_PROP_EXPOSURE, 156) #Set exposure value 1.0 - 5000 156.0#Red is selected by default. To recognize other colors, comment out the red range code below and uncomment the other color range code sections.# Red rangecolor_lower = np.array([0, 43, 46])color_upper = np.array([10, 255, 255])# #Green range# color_lower = np.array([35, 43, 46])# color_upper = np.array([77, 255, 255])# #Blue range# color_lower=np.array([100, 43, 46])# color_upper = np.array([124, 255, 255])# #Yellow range# color_lower = np.array([26, 43, 46])# color_upper = np.array([34, 255, 255])# #Orange range# color_lower = np.array([11, 43, 46])# color_upper = np.array([25, 255, 255])def Color_Recongnize():while(1):# get a frame and show Get the video frame and convert it to HSV format, use cvtColor() to convert the BGR format to HSV format, with the parameter cv2.COLOR_BGR2HSV.ret, frame = cap.read()#cv2.imshow('Capture', frame)origin_widget.value = bgr8_to_jpeg(frame)# change to hsv modelhsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)# get mask Use the inRange() function and the upper and lower bounds of the blue range in the HSV model to obtain the mask. In the mask, the blue part of the original video will be turned white, and the other parts black.mask = cv2.inRange(hsv, color_lower, color_upper)#cv2.imshow('Mask', mask)mask_widget.value = bgr8_to_jpeg(mask)# detect blue Perform a bitwise AND operation between the mask and the original video frame, which will replace the white in the mask with the real image:res = cv2.bitwise_and(frame, frame, mask=mask)#cv2.imshow('Result', res)result_widget.value = bgr8_to_jpeg(res)# if cv2.waitKey(1) & 0xFF == ord('q'):# breaktime.sleep(0.01)cap.release()#cv2.destroyAllWindows()#Start the processthread1 = threading.Thread(target=Color_Recongnize)thread1.setDaemon(True)thread1.start()#End the process, this code only needs to be executed when endingstop_thread(thread1)
This is the effect of performing red recognition. If you want to recognize other colors, you can comment out the red range and uncomment other color ranges.
For the CSI camera, the code path to execute is /home/pi/Yahboom_Project/1.OpenCV course/04 advanced tutorial/Color_recognition_CSI.ipynb
Note: For the CSI camera, if you want to re-run the program, you need to do the following:
