Aim: Recognition of captcha to find the Best accuracy of the text based captcha by using the Algorithms in Machine Learning. The two algorithms are K-Nearest Neighbour(KNN) and the Novel Support Vector Machine(SVM). Materials and Methods: The dataset is collected from www.kaggle.com. And the Two groups are K-Nearest Neighbour (N=10) and Novel Support Vector Machine(N=10) by using G-power and minimum power of the analysis is fixed as 80% and maximum accepted error is fixed as 0.5 with threshold value as 0.0805% and Confidence Interval is 95%. Results: The Novel Support vector machine is used to recognize the captcha. The accuracy found for the improved captcha is 99.98% and for the K-Nearest Neighbour is 98.78%. The two algorithms are used to find the improved classification or complexity of the captcha. The significant value obtained is p=0.002 (p<0.05) i.e α=0.05 and hence, there exists a statistically significant difference between the two groups with a confidence level of 95%. Conclusion: Recognizing the Captcha Recognition significantly seems to be better in Novel Support vector machines than k-nearest Neighbour.