Variables | |
| tuple | cfg = configuration.configuration('/home/martin/robot1_data/usr/martin/laser_camera_segmentation/calib') |
| print CvSVM::train(const CvMat* _train_data, const CvMat* _responses, const CvMat* _var_idx=0, const CvMat* _sample_idx=0, CvSVMParams _params=CvSVMParams()) train_data = cv.cvCreateMat(10,2,cv.CV_32FC1) #CvMat* cvCreateMat(int rows, int cols, int type) train_data[0][0] = 1 train_data[1][0] = 2 train_data[2][0] = 3 train_data[3][0] = 4 train_data[4][0] = 5 train_data[5][0] = 6 train_data[6][0] = 7 train_data[7][0] = 8 train_data[8][0] = 9 train_data[9][0] = 10 train_data[0][1] = 1 train_data[1][1] = 2 train_data[2][1] = 3 train_data[3][1] = 4 train_data[4][1] = 5 train_data[5][1] = 6 train_data[6][1] = 7 train_data[7][1] = 8 train_data[8][1] = 9 train_data[9][1] = 10 | |
| tuple | sc = scanner.scanner(cfg) |
| tuple laser_camera_segmentation::test::cfg = configuration.configuration('/home/martin/robot1_data/usr/martin/laser_camera_segmentation/calib') |
print CvSVM::train(const CvMat* _train_data, const CvMat* _responses, const CvMat* _var_idx=0, const CvMat* _sample_idx=0, CvSVMParams _params=CvSVMParams()) train_data = cv.cvCreateMat(10,2,cv.CV_32FC1) #CvMat* cvCreateMat(int rows, int cols, int type) train_data[0][0] = 1 train_data[1][0] = 2 train_data[2][0] = 3 train_data[3][0] = 4 train_data[4][0] = 5 train_data[5][0] = 6 train_data[6][0] = 7 train_data[7][0] = 8 train_data[8][0] = 9 train_data[9][0] = 10 train_data[0][1] = 1 train_data[1][1] = 2 train_data[2][1] = 3 train_data[3][1] = 4 train_data[4][1] = 5 train_data[5][1] = 6 train_data[6][1] = 7 train_data[7][1] = 8 train_data[8][1] = 9 train_data[9][1] = 10
for i in range(10): print train_data[i][0] print train_data[i][1] print '###'
responses = cv.cvCreateMat(10,1,cv.CV_32FC1) responses[0] = 1 responses[1] = 1 responses[2] = 1 responses[3] = 1 responses[4] = 1 responses[5] = 0 responses[6] = 0 responses[7] = 0 responses[8] = 0 responses[9] = 0
params = cv.CvSVMParams() params.svm_type = cv.CvSVM.C_SVC Type of SVM, one of the following types: CvSVM::C_SVC - n-class classification (n>=2), allows imperfect separation of classes with penalty multiplier C for outliers. CvSVM::NU_SVC - n-class classification with possible imperfect separation. Parameter nu (in the range 0..1, the larger the value, the smoother the decision boundary) is used instead of C. CvSVM::ONE_CLASS - one-class SVM. All the training data are from the same class, SVM builds a boundary that separates the class from the rest of the feature space. CvSVM::EPS_SVR - regression. The distance between feature vectors from the training set and the fitting hyperplane must be less than p. For outliers the penalty multiplier C is used. CvSVM::NU_SVR - regression; nu is used instead of p. params.kernel_type = cv.CvSVM.SIGMOID CvSVM::LINEAR - no mapping is done, linear discrimination (or regression) is done in the original feature space. It is the fastest option. d(x,y) = x*y == (x,y) CvSVM::POLY - polynomial kernel: d(x,y) = (gamma*(x*y)+coef0)degree CvSVM::RBF - radial-basis-function kernel; a good choice in most cases: d(x,y) = exp(-gamma*|x-y|2) CvSVM::SIGMOID - sigmoid function is used as a kernel: d(x,y) = tanh(gamma*(x*y)+coef0)
print my_svm.train_auto(train_data, responses,None,None,params) print my_svm.get_params() test = cv.cvCreateMat(1,2,cv.CV_32FC1) test[0] = 6 test[1] = 8.7878 print my_svm.predict(test)
import matplotlib.pyplot as plt import matplotlib.image as mpimg import numpy as np
n = 100 m = 100 results = np.array(-1*np.ones((n,m)))
for i in range(n): for j in range(m): test[0]=i test[1]=j results[i][j] = my_svm.predict(test) print str(i) + ' ' + str(j) + ' ' + ' -> ' + str(results[i][j])
print results
imgplot = plt.imshow(results, cmap=pylab.cm.gray, interpolation='nearest') imgplot = plt.imshow(np.array(train_data).transpose()) imgscatter = plt.scatter(np.array(train_data)[:,0], np.array(train_data)[:,1]) plt.show()
pylab.ion() #interactive pylab.figure(figsize=(8,4)) pylab.hold(True) pylab.subplot(121) pylab.title('test') pylab.imshow(responses, cmap=pylab.cm.gray, interpolation='nearest')
pylab.draw()
cfg = configuration.configuration('/home/martin/robot1_data/usr/martin/laser_camera_segmentation/labeling') sc = scanner.scanner(cfg) pc = processor.processor(cfg)
name = ut.formatted_time() sc.capture_and_save(name) pc.load_raw_data(name)
id = '2009Sep14_095609' pc.load_raw_data(id) pc.load_metadata(id) print pc.scan_dataset.id print pc.scan_dataset.polygons pc.create_polygon_images() pc.process_raw_data() pc.save_mapped_image(name) pc.display_all_data()
print pc.scan_dataset.polygons[0].cvImage[400]
! /usr/bin/env python
Definition at line 171 of file laser_camera_segmentation/test.py.
Definition at line 173 of file laser_camera_segmentation/test.py.