How to count images

vision
1

How can I count the agave plants in this field? what technique do you recommend??
https://drive.google.com/file/d/1UWf1p2juWVudpG7-I2oVI2B-M4xO3Izz/view?usp=sharing

thanks

2

You could start with a naive approach of trying to predict the target value directly or alternatively use an object detection task, where you could count the predicted bounding boxes for each plant.

3

Captura de pantalla 2020-06-30 a la(s) 9.32.30

as you see not being able to identify only the plants:
my code is


#conversion a escala de grises
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
gray = cv2.GaussianBlur(gray,(25,25),0)


ret, threshed_img = cv2.threshold(cv2.cvtColor(img, cv2.COLOR_BGR2GRAY),
                127, 255, cv2.THRESH_BINARY)
# find contours and get the external one

dilated = cv2.dilate(thresh,kernel,iterations = 13) # dilate
contours, hierarchy = cv2.findContours(dilated,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_NONE)

for cnt in contours:
    x,y,w,h = cv2.boundingRect(cnt)
    if h<30 or w<30:
        continue
    else:
        
        cv2.rectangle(img, (x, y), (x+w, y+h), (0, 255, 0), 2)
            # get the min area rect
        rect = cv2.minAreaRect(cnt)
        box = cv2.boxPoints(rect)
        # convert all coordinates floating point values to int
        box = np.int0(box)
        # draw a red 'nghien' rectangle
        #cv2.drawContours(img, [box], 0, (0, 0, 255))
    
        # finally, get the min enclosing circle
        (x, y), radius = cv2.minEnclosingCircle(cnt)
        # convert all values to int
        center = (int(x), int(y))
        radius = int(radius)
        # and draw the circle in blue
        img = cv2.circle(img, center, radius, (255, 0, 0), 2)
    
    

print(len(contours))
4

If you don’t have the labels and would like to use a classical computer vision approach, I would recommend to post the question in an OpenCV-specific discussion board or on StackOverflow, as your use case doesn’t seem to be PyTorch-specific. :wink: