from startup import * from ijutils import * from ch.psi.pshell.imaging.Overlays import * import ch.psi.pshell.imaging.Pen as Pen import java.awt.Color as Color import random import ch.psi.pshell.imaging.ImageListener as ImageListener from operator import add, mul, sub, truediv from ch.psi.pshell.imaging.Utils import sub def arrmul(a, b): """Multiply 2 series of the same size. Args: a(list, tuple, array ...): subscriptable object containing numbers b(list, tuple, array ...): subscriptable object containing numbers Returns: List """ return map(mul, a, b) def center_of_mass(data, x = None): """Calculate the center of mass of a series, and its rms. Args: data(list, tuple, array ...): subscriptable object containing numbers x(list, tuple, array ..., optional): x coordinates Returns: Tuple (com, rms) """ if x is None: x = Arr.indexesDouble(len(data)) data_sum = sum(data) if (data_sum==0): return float('nan') xmd = arrmul( x, data) com = sum(xmd) / data_sum xmd2 = arrmul( x, xmd) com2 = sum(xmd2) / data_sum rms = math.sqrt(abs(com2 - com * com)) return (com, rms) def get_centroid(source): bi = source.getImage() if bi is None: return None op = show_panel(bi, "Original") ip = load_image(bi) plot(get_histogram(ip), title = "Histogram") grayscale(ip) invert(ip) gaussian_blur(ip) auto_threshold(ip) #binary_erode(ip) show_panel(ip.getBufferedImage(), "Image") (results,output_img)=analyse_particles(ip, 2000,20000, exclude_edges=False, print_table=True) op.clearOverlays() show_panel(output_img.getBufferedImage(), "Outlines") if results.size()>0: centroid = (results.getValue("XM",0), results.getValue("YM",0)) ov = Crosshairs(Pen(Color.ORANGE), java.awt.Point(int(centroid[0]),int(centroid[1])), java.awt.Dimension(15,15)) op.addOverlay(ov) return centroid import ch.psi.pshell.imaging.Filter as Filter class SimulatedSource(Filter): def process(self, img, data): self.img=img if img is None: return None ip = load_image(img) pad_h = int((random.random()-0.5) * 500) pad_v = int((random.random()-0.5) * 500) #print "Pad = " , (pad_h, pad_v) ip = pad_image(ip, -pad_h, pad_h, pad_v, -pad_v, fill_color = Color.BLACK) return ip.getBufferedImage() #return img def waitNext(self, timeout): self.pushImage(self.process(self.img, None)) class ImageStats(DeviceBase): def __init__(self, name, source): DeviceBase.__init__(self, name) self.source = source self.com_x_samples, self.com_y_samples = [], [] self.rms_x_samples, self.rms_y_samples = [], [] class ComX(Readable): def read(self): if len(self.image_stats.com_x_samples)==0: return None return mean(self.image_stats.com_x_samples) self.com_x_mean = ComX(); self.com_x_mean.image_stats = self class ComY(Readable): def read(self): if len(self.image_stats.com_y_samples)==0: return None return mean(self.image_stats.com_y_samples) self.com_y_mean = ComY(); self.com_y_mean.image_stats = self class ComXVar(Readable): def read(self): if len(self.image_stats.com_x_samples)==0: return None return stdev(self.image_stats.com_x_samples) self.com_x_stdev = ComXVar(); self.com_x_stdev.image_stats = self class ComYVar(Readable): def read(self): if len(self.image_stats.com_y_samples)==0: return None return stdev(self.image_stats.com_y_samples) self.com_y_stdev = ComYVar(); self.com_y_stdev.image_stats = self set_device_alias(self.com_x_mean, name + " com x mean") set_device_alias(self.com_y_mean, name + " com y mean") set_device_alias(self.com_x_stdev, name + " com x stdev") set_device_alias(self.com_y_stdev, name + " com y stdev") self.bg_en = False self.num_images = 5 self.initialize() #class SourceListener (ImageListener): # def __init__(self, dev): # self.dev=dev # def onImage(self, origin, image, data): # self.dev.doUpdate() # def onError(self, origin, ex): # self.dev.com_x_samples, self.dev.com_y_samples = [], [] # self.rms_x_samples, self.rms_y_samples = [], [] #self.listener = SourceListener(self) #self.source.addListener(self.listener) def doUpdate(self): #print "Do update" self.com_x_samples, self.com_y_samples = [], [] self.rms_x_samples, self.rms_y_samples = [], [] for i in range(self.num_images): if type(self.source) is not ch.psi.pshell.imaging.FileSource: self.source.waitNext(5000) #time.sleep(0.2) #print "Ok" #centroid = get_centroid(self.source) #print "cent ", centroid #if centroid is not None: # self.com_x_samples.append(centroid[0]) # self.com_y_samples.append(centroid[1]) x_profile = self.source.data.integrateVertically(True) y_profile = self.source.data.integrateHorizontally(True) com_x,rms_x = center_of_mass(x_profile, self.source.data.getRowSelectionX(True)) com_y,rms_y = center_of_mass(y_profile, self.source.data.getColSelectionX(True)) self.com_x_samples.append(com_x) self.com_y_samples.append(com_y) self.rms_x_samples.append(rms_x) self.rms_y_samples.append(rms_y) def setNumberOfImages(self, value): self.num_images = value def enableBackground(self, value): self.bg_en = value class BackgroundSubtractor(Filter): def __init__(self, obj): self.obj=obj def process(self, image, data): return sub(image, obj.background, True) if (self.obj.background is not None) else None self.source.filter = BackgroundSubtractor(self) if (value==True) else None def captureBackground(self, images): self.doInitialize() def doClose(self): print "close" self.source.filter = None #self.source.removeListener(self.listener) def start(self): pass def stop(self): pass def get_simulated_source(img): simulated_source = SimulatedSource() simulated_source.img=None simulated_source.initialize() img.addListener(simulated_source) show_panel(simulated_source) return simulated_source if __name__ == "__builtin__": #simulated_source = get_simulated_source(image) #print get_centroid(simulated_source) add_device(ImageStats("image_stats", cam3), True) #cam3.waitNext(5000) image_stats.enableBackground(False) #for i in range (10): # image_stats.update() # print image_stats.take(), image_stats.com_x_mean.read(), image_stats.com_y_mean.read() # time.sleep(1) image_stats.setNumberOfImages(3) sensors = [image_stats.com_x_mean, image_stats.com_y_mean, image_stats.com_x_stdev, image_stats.com_y_stdev] def before_sample(): image_stats.update() tscan(sensors, 10, 0.1, before_read = before_sample)